| {"task_id": "data-42-0000", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,57\n1,65\n2,73\n3,-79\n4,89\n5,97\n6,105\n7,113\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "3", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0001", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,125\n1,127\n2,129\n3,131\n4,133\n5,135\n6,177\n7,139\n8,141\n9,143\n10,145\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0002", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,23,Madrid,hr,121000\n2,olive,60,London,finance,135000\n3,frank,61,London,sales,64000\n4,eve,33,Cairo,hr,118000\n5,bob,23,Berlin,ops,101000\n6,heidi,50,London,hr,71000\n7,karl,59,Tokyo,hr,50000\n8,nora,60,Paris,marketing,86000\n9,ivan,38,London,ops,142000\n10,carol,23,Madrid,eng,90000\n11,judy,29,London,ops,54000\n12,alice,21,Tokyo,hr,116000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "480", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0003", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,dave,49,London,eng,104000\n2,grace,54,Berlin,ops,70000\n3,heidi,26,Madrid,eng,58000\n4,pete,42,London,marketing,100000\n5,ivan,36,Lisbon,marketing,54000\n6,eve,45,Tokyo,finance,92000\n7,carol,46,Berlin,ops,124000\n8,linda,59,Madrid,hr,141000\n9,bob,55,Rome,sales,102000\n10,alice,41,Madrid,sales,94000\n11,mike,30,Paris,finance,106000\n12,olive,33,Berlin,finance,147000\n13,karl,20,Tokyo,hr,91000\n\n=== bonus_by_dept ===\ndept,bonus\neng,6000\nfinance,9000\nhr,4000\nmarketing,7000\nops,4000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1368000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0004", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,62,Cairo,sales,68000\n2,judy,21,Cairo,marketing,135000\n3,dave,51,Tokyo,finance,89000\n4,frank,33,Rome,ops,40000\n5,alice,32,Lisbon,marketing,114000\n6,mike,20,London,hr,83000\n7,linda,33,Cairo,finance,124000\n8,carol,30,Madrid,marketing,90000\n9,bob,60,Lisbon,ops,70000\n10,pete,30,Lisbon,hr,59000\n11,grace,26,Cairo,hr,56000\n12,eve,27,Cairo,finance,130000\n13,heidi,64,Berlin,hr,83000\n14,ivan,49,Rome,finance,123000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1264000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0005", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,53,Cairo,marketing,67000\n2,pete,48,Berlin,hr,104000\n3,linda,39,Paris,sales,40000\n4,eve,62,London,marketing,82000\n5,nora,31,Berlin,hr,144000\n6,heidi,35,London,sales,62000\n7,karl,28,Paris,ops,136000\n8,mike,56,Berlin,hr,66000\n9,dave,39,London,marketing,121000\n10,alice,42,Cairo,ops,125000\n11,grace,38,Cairo,hr,128000\n12,frank,63,Berlin,marketing,86000\n13,ivan,25,London,marketing,59000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "marketing", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0006", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,67\n1,71\n2,75\n3,79\n4,83\n5,87\n6,91\n7,95\n8,99\n9,23\n10,107\n11,111\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "9", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0007", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,56,Tokyo,finance,78000\n2,bob,32,Madrid,sales,146000\n3,linda,49,London,ops,47000\n4,karl,22,Paris,ops,47000\n5,judy,62,Madrid,marketing,52000\n6,frank,64,Paris,finance,50000\n7,mike,29,Madrid,hr,86000\n8,pete,26,Berlin,eng,59000\n9,grace,44,Tokyo,finance,75000\n10,nora,52,Tokyo,ops,79000\n11,olive,51,Paris,hr,84000\n12,alice,51,Berlin,marketing,91000\n13,eve,42,Tokyo,sales,94000\n14,dave,51,Rome,hr,69000\n15,ivan,31,Lisbon,marketing,59000\n```\n\nTask: Group the rows by 'dept' and compute the max of 'age' per group. Return the group name with the largest max.", "reference_answer": "finance", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0008", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,168\n1,173\n2,178\n3,183\n4,188\n5,193\n6,198\n7,203\n8,208\n9,213\n10,218\n11,323\n12,228\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "11", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0009", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,alice,22,London,ops,46000\n2,judy,30,Rome,sales,110000\n3,frank,47,London,ops,121000\n4,carol,24,Tokyo,finance,65000\n5,mike,49,London,hr,65000\n6,grace,57,London,eng,129000\n7,karl,27,Paris,sales,103000\n8,nora,45,Madrid,hr,112000\n9,heidi,51,Tokyo,ops,123000\n10,olive,21,Cairo,finance,93000\n\n=== bonus_by_dept ===\ndept,bonus\neng,8000\nfinance,9000\nhr,2000\nops,2000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1005000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0010", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,119\n1,122\n2,125\n3,128\n4,131\n5,134\n6,137\n7,140\n8,83\n9,146\n10,149\n11,152\n12,155\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "8", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0011", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,22,Cairo,hr,125000\n2,karl,41,Lisbon,finance,125000\n3,alice,33,Paris,ops,98000\n4,pete,57,Tokyo,ops,119000\n5,bob,48,Cairo,ops,111000\n6,eve,34,London,eng,45000\n7,dave,58,Berlin,marketing,75000\n8,linda,40,Rome,eng,49000\n9,judy,47,Paris,marketing,55000\n10,ivan,21,Lisbon,sales,72000\n11,grace,56,Madrid,marketing,142000\n12,nora,20,Cairo,hr,145000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1161000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0012", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,47,London,finance,87000\n2,grace,26,London,finance,135000\n3,karl,20,Paris,sales,49000\n4,eve,40,Rome,finance,98000\n5,nora,28,Tokyo,ops,120000\n6,ivan,39,Rome,eng,53000\n7,carol,32,Berlin,ops,103000\n8,olive,32,Rome,ops,45000\n9,bob,28,Madrid,eng,99000\n10,mike,45,Rome,eng,139000\n11,pete,37,Paris,sales,88000\n12,alice,39,Rome,marketing,137000\n13,heidi,31,Cairo,sales,128000\n14,linda,46,Cairo,ops,91000\n15,frank,26,Tokyo,eng,148000\n16,dave,63,Tokyo,ops,75000\n17,judy2,28,Paris,ops,100000\n18,grace2,63,Lisbon,ops,140000\n19,karl2,45,Tokyo,marketing,47000\n20,eve2,58,Rome,marketing,137000\n21,nora2,46,Paris,marketing,72000\n22,ivan2,55,Tokyo,marketing,129000\n23,carol2,39,London,marketing,124000\n```\n\nTask: How many rows have age > 31 AND city = 'Tokyo'? Return the integer count.", "reference_answer": "3", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0013", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,ivan,34,Lisbon,finance,52000\n2,bob,23,Cairo,sales,132000\n3,heidi,25,Berlin,ops,61000\n4,grace,62,Lisbon,finance,113000\n5,linda,49,Paris,ops,40000\n6,mike,39,Tokyo,sales,100000\n7,dave,40,Paris,hr,127000\n8,pete,52,Rome,marketing,130000\n9,nora,54,Tokyo,eng,100000\n10,judy,58,Tokyo,hr,81000\n11,frank,60,Paris,eng,115000\n\n=== bonus_by_dept ===\ndept,bonus\neng,7000\nfinance,7000\nhr,8000\nmarketing,6000\nops,7000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1131000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0014", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,judy,63,Tokyo,eng,48000\n2,pete,37,Tokyo,hr,130000\n3,carol,55,Cairo,hr,94000\n4,grace,43,Lisbon,sales,118000\n5,mike,44,Tokyo,ops,63000\n6,nora,43,Paris,hr,65000\n7,dave,62,Lisbon,eng,72000\n8,frank,43,Lisbon,marketing,147000\n9,eve,36,Tokyo,sales,138000\n10,olive,45,Madrid,sales,75000\n11,heidi,53,Berlin,ops,74000\n\n=== bonus_by_dept ===\ndept,bonus\neng,9000\nhr,3000\nmarketing,1000\nops,3000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1082000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0015", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,47,Rome,hr,74000\n2,heidi,63,Madrid,ops,99000\n3,mike,48,Rome,finance,54000\n4,linda,60,Cairo,eng,70000\n5,judy,26,Cairo,ops,83000\n6,grace,34,Tokyo,eng,145000\n7,pete,26,Cairo,sales,116000\n8,eve,45,Rome,eng,92000\n```\n\nTask: Compute the max of the 'age' column. Return the integer.", "reference_answer": "63", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0016", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,127\n1,131\n2,135\n3,139\n4,143\n5,147\n6,231\n7,155\n8,159\n9,163\n10,167\n11,171\n12,175\n13,179\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0017", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,42,Madrid,eng,106000\n2,nora,31,Tokyo,hr,63000\n3,ivan,26,Cairo,ops,47000\n4,dave,41,Cairo,eng,44000\n5,olive,47,Lisbon,ops,144000\n6,eve,22,Lisbon,hr,134000\n7,pete,43,Madrid,eng,108000\n8,alice,29,Berlin,ops,105000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 1 (0-indexed, header excluded)? Return only the value.", "reference_answer": "hr", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0018", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,28,Lisbon,ops,50000\n2,ivan,46,London,finance,98000\n3,grace,31,Berlin,ops,74000\n4,karl,52,London,marketing,132000\n5,heidi,26,Lisbon,ops,82000\n6,alice,64,Rome,marketing,75000\n7,eve,64,London,hr,109000\n8,bob,42,Rome,ops,47000\n9,pete,54,Lisbon,marketing,116000\n10,nora,48,Tokyo,eng,102000\n11,dave,58,Rome,eng,71000\n12,olive,49,Berlin,marketing,63000\n13,frank,40,Berlin,eng,51000\n14,carol,38,Paris,eng,96000\n15,linda,20,Berlin,finance,115000\n16,mike,50,Cairo,finance,127000\n17,judy2,25,Madrid,finance,95000\n```\n\nTask: How many rows have age > 29 AND city = 'Madrid'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0019", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,82\n1,91\n2,100\n3,109\n4,118\n5,127\n6,136\n7,145\n8,154\n9,-17\n10,172\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "9", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0020", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,27,Lisbon,marketing,137000\n2,alice,45,Rome,marketing,84000\n3,olive,31,Madrid,hr,42000\n4,pete,43,Lisbon,finance,147000\n5,mike,43,Cairo,ops,43000\n6,karl,63,Madrid,finance,50000\n7,bob,39,Berlin,ops,67000\n8,ivan,49,Paris,ops,42000\n9,carol,31,Lisbon,sales,59000\n10,eve,61,London,ops,114000\n11,frank,22,Tokyo,ops,40000\n12,dave,44,Madrid,eng,77000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "Lisbon", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0021", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,21,Berlin,eng,139000\n2,nora,43,Madrid,marketing,143000\n3,karl,33,Madrid,marketing,68000\n4,mike,55,Madrid,eng,66000\n5,dave,58,Madrid,finance,60000\n6,linda,28,Tokyo,marketing,80000\n7,olive,51,Madrid,eng,114000\n8,pete,24,Tokyo,marketing,121000\n9,eve,55,Paris,ops,124000\n10,heidi,32,Paris,ops,100000\n11,judy,63,Berlin,finance,82000\n12,grace,40,Rome,sales,113000\n13,alice,56,Berlin,eng,54000\n14,bob,55,London,eng,72000\n15,carol,37,Paris,hr,123000\n16,ivan,29,Paris,eng,107000\n```\n\nTask: How many rows have age > 27 AND city = 'London'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0022", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,59,Rome,hr,105000\n2,grace,28,Tokyo,finance,101000\n3,carol,52,Paris,ops,120000\n4,nora,25,Rome,hr,53000\n5,frank,29,Madrid,finance,66000\n6,linda,62,Berlin,marketing,122000\n7,bob,47,London,sales,114000\n8,karl,63,Tokyo,finance,137000\n9,pete,38,Paris,ops,72000\n10,olive,51,Lisbon,ops,122000\n11,heidi,45,Paris,sales,56000\n12,alice,26,London,marketing,128000\n13,mike,29,Rome,finance,45000\n14,dave,36,Madrid,eng,129000\n15,eve,56,Paris,marketing,140000\n16,ivan,30,Cairo,ops,48000\n```\n\nTask: Compute the max of the 'age' column. Return the integer.", "reference_answer": "63", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0023", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,173\n1,178\n2,183\n3,188\n4,193\n5,198\n6,203\n7,208\n8,213\n9,218\n10,123\n11,228\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "10", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0024", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,63,Madrid,finance,99000\n2,pete,41,Rome,marketing,86000\n3,dave,27,Lisbon,ops,107000\n4,nora,50,London,sales,136000\n5,karl,52,Rome,ops,126000\n6,carol,40,London,finance,90000\n7,linda,42,Berlin,finance,142000\n8,mike,23,London,finance,113000\n9,judy,59,Rome,marketing,55000\n10,olive,52,Rome,eng,120000\n11,ivan,60,Rome,marketing,139000\n12,grace,31,London,eng,64000\n13,alice,21,London,sales,69000\n14,bob,29,Lisbon,hr,89000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1435000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0025", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,60,Paris,marketing,128000\n2,pete,27,Madrid,ops,93000\n3,ivan,57,Madrid,sales,133000\n4,nora,26,Berlin,eng,130000\n5,judy,26,Rome,sales,148000\n6,mike,45,Tokyo,eng,53000\n7,eve,63,Lisbon,marketing,116000\n8,heidi,29,Tokyo,sales,76000\n9,alice,41,Rome,finance,64000\n10,carol,55,London,hr,96000\n11,frank,35,Rome,finance,116000\n12,bob,47,Cairo,marketing,83000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "Lisbon", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0026", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,alice,28,London,marketing,114000\n2,grace,61,Cairo,sales,135000\n3,heidi,46,Paris,ops,85000\n4,eve,43,Paris,finance,105000\n5,linda,42,Tokyo,finance,52000\n6,dave,38,Cairo,marketing,126000\n7,mike,25,London,finance,45000\n8,carol,36,Rome,ops,76000\n9,olive,21,Cairo,ops,96000\n10,nora,52,Cairo,marketing,98000\n11,frank,40,Madrid,hr,106000\n12,ivan,41,Lisbon,sales,90000\n13,bob,38,Cairo,marketing,42000\n14,pete,31,Paris,finance,135000\n```\n\nTask: Group the rows by 'city' and compute the max of 'salary' per group. Return the group name with the largest max.", "reference_answer": "Cairo", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0027", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,bob,49,Paris,ops,59000\n2,karl,41,Cairo,sales,40000\n3,frank,20,Tokyo,eng,116000\n4,judy,63,Paris,marketing,81000\n5,alice,37,Rome,hr,52000\n6,heidi,44,Madrid,ops,41000\n7,dave,62,Berlin,finance,70000\n8,carol,44,Tokyo,sales,142000\n9,olive,57,Tokyo,ops,100000\n10,grace,53,Rome,sales,42000\n11,pete,58,Rome,sales,75000\n12,ivan,49,Paris,sales,105000\n13,eve,27,Berlin,finance,117000\n14,linda,32,Lisbon,hr,41000\n15,nora,31,Lisbon,sales,91000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1172000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0028", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,judy,40,London,hr,126000\n2,alice,20,Berlin,marketing,87000\n3,frank,57,Lisbon,eng,47000\n4,mike,60,Madrid,hr,40000\n5,dave,30,Rome,eng,136000\n6,heidi,37,Rome,eng,124000\n7,olive,44,Madrid,marketing,81000\n8,ivan,25,Rome,marketing,44000\n\n=== bonus_by_dept ===\ndept,bonus\neng,5000\nhr,3000\nmarketing,2000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "712000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0029", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,134\n1,142\n2,-10\n3,158\n4,166\n5,174\n6,182\n7,190\n8,198\n9,206\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "2", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0030", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,38,Paris,sales,139000\n2,olive,24,Rome,marketing,106000\n3,grace,25,Berlin,eng,77000\n4,karl,39,Madrid,eng,55000\n5,carol,20,Lisbon,ops,75000\n6,heidi,38,Paris,finance,114000\n7,pete,50,Madrid,eng,54000\n8,nora,51,Berlin,ops,60000\n9,dave,27,Madrid,sales,135000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 1 (0-indexed, header excluded)? Return only the value.", "reference_answer": "Rome", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0031", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,38,Cairo,ops,139000\n2,alice,54,Paris,eng,105000\n3,bob,26,Rome,finance,51000\n4,grace,21,Tokyo,marketing,45000\n5,frank,22,Paris,hr,69000\n6,linda,23,Cairo,eng,43000\n7,judy,53,Lisbon,finance,60000\n8,heidi,46,Paris,hr,102000\n```\n\nTask: Compute the median of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "64500.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0032", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,99\n1,101\n2,103\n3,105\n4,107\n5,109\n6,111\n7,113\n8,75\n9,117\n10,119\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "8", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0033", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,mike,58,Cairo,ops,111000\n2,linda,42,Madrid,marketing,46000\n3,ivan,53,Rome,eng,135000\n4,grace,33,London,hr,119000\n5,alice,55,Paris,finance,67000\n6,judy,54,London,marketing,94000\n7,dave,54,London,hr,87000\n8,frank,42,Rome,sales,130000\n9,nora,50,London,marketing,142000\n\n=== bonus_by_dept ===\ndept,bonus\neng,6000\nfinance,9000\nhr,5000\nmarketing,4000\nops,7000\nsales,9000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "984000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0034", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,62,Cairo,hr,49000\n2,nora,62,Lisbon,hr,60000\n3,linda,31,Lisbon,ops,123000\n4,mike,50,Tokyo,ops,146000\n5,karl,46,Rome,hr,135000\n6,olive,43,Tokyo,finance,74000\n7,bob,56,Lisbon,eng,140000\n8,ivan,28,Cairo,ops,120000\n9,carol,22,Cairo,ops,83000\n10,pete,29,London,ops,110000\n11,judy,40,London,ops,51000\n12,eve,45,Madrid,finance,55000\n13,alice,43,Rome,finance,102000\n14,grace,31,Berlin,eng,122000\n15,frank,43,Rome,eng,121000\n16,heidi,24,Cairo,eng,63000\n17,dave2,34,Cairo,eng,53000\n```\n\nTask: How many rows have age > 35 AND city = 'Cairo'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0035", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,75\n1,201\n2,87\n3,93\n4,99\n5,105\n6,111\n7,117\n8,123\n9,129\n10,135\n11,141\n12,147\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "1", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0036", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,53,Berlin,hr,109000\n2,eve,24,Lisbon,sales,135000\n3,nora,26,Berlin,marketing,143000\n4,judy,50,Lisbon,eng,51000\n5,frank,20,Berlin,sales,56000\n6,dave,20,Cairo,ops,72000\n7,heidi,59,Cairo,eng,110000\n8,karl,35,Lisbon,hr,135000\n9,mike,36,Tokyo,ops,147000\n10,olive,45,Rome,ops,54000\n11,alice,49,Lisbon,sales,122000\n12,ivan,38,Cairo,hr,79000\n13,grace,52,Madrid,ops,142000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "38.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0037", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,61,Berlin,marketing,135000\n2,heidi,48,Cairo,eng,42000\n3,bob,46,Madrid,sales,64000\n4,karl,50,Paris,marketing,83000\n5,ivan,53,Paris,eng,63000\n6,judy,26,Rome,hr,138000\n7,mike,34,Berlin,ops,132000\n8,linda,57,Berlin,finance,98000\n```\n\nTask: In the CSV below, what is the value of column 'salary' in row 6 (0-indexed, header excluded)? Return only the value.", "reference_answer": "132000", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0038", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,62,Paris,ops,56000\n2,judy,45,London,hr,82000\n3,ivan,59,Paris,hr,93000\n4,mike,26,Rome,finance,81000\n5,carol,62,Paris,ops,119000\n6,olive,52,London,hr,57000\n7,grace,21,Madrid,marketing,125000\n8,heidi,25,Lisbon,finance,121000\n9,pete,40,Berlin,eng,73000\n10,nora,46,Tokyo,marketing,71000\n11,dave,38,Lisbon,finance,99000\n12,alice,62,Paris,sales,56000\n13,eve,58,Tokyo,sales,102000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1135000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0039", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,50,Cairo,hr,59000\n2,carol,42,Lisbon,ops,125000\n3,ivan,35,Cairo,finance,145000\n4,heidi,63,Madrid,eng,124000\n5,frank,30,Madrid,hr,144000\n6,nora,46,Madrid,sales,51000\n7,eve,22,Lisbon,ops,40000\n8,karl,25,Rome,marketing,73000\n9,dave,47,Tokyo,hr,139000\n10,bob,56,Cairo,sales,79000\n11,linda,50,Cairo,finance,82000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 0 (0-indexed, header excluded)? Return only the value.", "reference_answer": "Cairo", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0040", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,25,Rome,ops,122000\n2,linda,52,Rome,eng,52000\n3,grace,22,Berlin,hr,129000\n4,alice,48,Tokyo,eng,46000\n5,carol,50,Paris,hr,109000\n6,ivan,56,Madrid,marketing,44000\n7,bob,61,Rome,finance,132000\n8,heidi,43,London,sales,90000\n9,olive,25,Cairo,marketing,147000\n10,mike,40,Berlin,finance,61000\n11,karl,21,London,hr,126000\n12,frank,64,Madrid,eng,101000\n13,pete,53,Rome,sales,48000\n14,dave,62,Rome,sales,134000\n15,nora,63,Tokyo,marketing,139000\n16,eve,64,Rome,marketing,98000\n```\n\nTask: How many rows have age > 25 AND city = 'Berlin'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0041", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,48,Paris,ops,107000\n2,carol,24,Cairo,finance,50000\n3,frank,63,Berlin,hr,98000\n4,bob,35,Madrid,marketing,106000\n5,ivan,64,Berlin,eng,77000\n6,nora,38,London,eng,142000\n7,linda,42,Cairo,ops,129000\n8,karl,42,Berlin,finance,70000\n9,judy,42,Rome,marketing,63000\n10,alice,25,Madrid,marketing,114000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 7 (0-indexed, header excluded)? Return only the value.", "reference_answer": "42", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0042", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,43,London,eng,86000\n2,judy,28,Rome,marketing,140000\n3,ivan,23,Lisbon,finance,55000\n4,karl,55,London,hr,145000\n5,frank,48,Cairo,ops,92000\n6,heidi,42,Lisbon,finance,67000\n7,carol,26,Cairo,sales,40000\n8,eve,60,Paris,hr,54000\n9,dave,51,Tokyo,ops,73000\n10,linda,63,Tokyo,ops,145000\n11,bob,54,Madrid,sales,104000\n12,grace,27,London,finance,101000\n13,mike,47,Berlin,eng,68000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "47.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0043", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,60,Berlin,finance,95000\n2,mike,52,Madrid,ops,112000\n3,judy,20,Berlin,sales,124000\n4,pete,25,Paris,sales,41000\n5,alice,34,Lisbon,hr,59000\n6,karl,57,Rome,eng,116000\n7,bob,47,Cairo,ops,55000\n8,heidi,55,Cairo,ops,95000\n9,linda,30,Berlin,marketing,110000\n10,frank,23,Madrid,eng,114000\n11,nora,39,Paris,hr,95000\n12,ivan,40,Madrid,sales,142000\n13,eve,23,Cairo,marketing,44000\n14,grace,35,Madrid,eng,56000\n15,carol,64,Cairo,ops,139000\n16,dave,63,Rome,marketing,117000\n17,olive2,30,Madrid,finance,134000\n18,mike2,42,Tokyo,finance,100000\n19,judy2,20,Cairo,eng,146000\n```\n\nTask: How many rows have age > 31 AND city = 'Madrid'? Return the integer count.", "reference_answer": "3", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0044", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,34,Paris,hr,111000\n2,bob,21,Lisbon,hr,142000\n3,pete,44,Lisbon,finance,145000\n4,eve,28,Lisbon,ops,80000\n5,alice,45,Tokyo,sales,90000\n6,linda,44,Lisbon,marketing,145000\n7,mike,37,London,ops,148000\n8,judy,59,Madrid,finance,104000\n9,grace,31,Berlin,eng,53000\n10,karl,36,Paris,sales,42000\n11,dave,35,Tokyo,hr,59000\n12,olive,58,Paris,ops,59000\n13,nora,33,Rome,eng,108000\n14,ivan,34,Madrid,ops,136000\n15,carol,24,Paris,finance,108000\n16,frank,26,Rome,sales,149000\n17,heidi2,43,Paris,marketing,52000\n```\n\nTask: How many rows have age > 38 AND city = 'Rome'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0045", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,45,Lisbon,finance,139000\n2,bob,52,Cairo,marketing,48000\n3,frank,44,Berlin,hr,81000\n4,alice,45,Paris,hr,48000\n5,judy,43,Cairo,hr,135000\n6,linda,62,Madrid,marketing,96000\n7,karl,62,Rome,eng,145000\n8,carol,41,Berlin,sales,99000\n9,dave,21,London,marketing,60000\n10,mike,64,Cairo,sales,107000\n11,grace,27,Berlin,eng,63000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "506", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0046", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,alice,30,Paris,hr,72000\n2,dave,43,Cairo,eng,56000\n3,olive,61,Berlin,eng,122000\n4,grace,48,Berlin,eng,116000\n5,nora,41,Lisbon,sales,52000\n6,heidi,29,Lisbon,finance,107000\n7,carol,48,Madrid,hr,54000\n8,pete,33,Cairo,sales,143000\n9,frank,37,Cairo,ops,49000\n10,linda,48,Tokyo,hr,40000\n11,eve,64,Cairo,hr,118000\n12,mike,26,Rome,hr,96000\n13,judy,41,Tokyo,finance,71000\n14,ivan,38,Lisbon,sales,41000\n```\n\nTask: Compute the mean of the 'age' column. Return the numeric value (any precision).", "reference_answer": "41.92857142857143", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0047", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,51,Paris,marketing,63000\n2,alice,29,Berlin,marketing,110000\n3,frank,39,Rome,hr,147000\n4,linda,43,Paris,finance,124000\n5,eve,45,Paris,marketing,136000\n6,olive,50,Tokyo,sales,117000\n7,pete,25,Madrid,sales,127000\n8,heidi,43,Berlin,ops,43000\n9,bob,47,London,eng,112000\n10,nora,43,Paris,sales,142000\n11,mike,36,London,hr,145000\n12,carol,51,Cairo,eng,55000\n13,dave,48,Paris,ops,140000\n14,judy,44,Lisbon,marketing,66000\n15,grace,51,London,hr,128000\n16,ivan,37,Madrid,ops,112000\n17,karl2,51,Madrid,ops,41000\n18,alice2,58,London,ops,53000\n19,frank2,31,Madrid,sales,124000\n20,linda2,33,Tokyo,hr,72000\n```\n\nTask: How many rows have age > 42 AND city = 'Rome'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0048", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,37,Berlin,hr,99000\n2,olive,40,Rome,hr,61000\n3,frank,53,Paris,finance,108000\n4,bob,43,Paris,sales,147000\n5,grace,43,Rome,sales,108000\n6,nora,20,Madrid,ops,51000\n7,linda,33,Madrid,hr,118000\n8,dave,53,Madrid,sales,82000\n9,pete,61,Cairo,hr,146000\n10,carol,28,London,eng,140000\n11,eve,49,Lisbon,marketing,116000\n12,alice,27,Madrid,finance,117000\n```\n\nTask: Compute the max of the 'age' column. Return the integer.", "reference_answer": "61", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0049", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,45,Paris,eng,126000\n2,nora,20,Cairo,ops,65000\n3,olive,59,Lisbon,finance,55000\n4,eve,49,London,eng,148000\n5,dave,33,Cairo,marketing,135000\n6,grace,36,Madrid,ops,145000\n7,heidi,33,Madrid,finance,149000\n8,frank,43,Berlin,finance,110000\n9,alice,44,Rome,eng,92000\n10,mike,62,Paris,finance,51000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "149000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0050", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,156\n1,162\n2,168\n3,174\n4,180\n5,66\n6,192\n7,198\n8,204\n9,210\n10,216\n11,222\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "5", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0051", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,judy,48,Berlin,hr,89000\n2,heidi,53,Berlin,ops,108000\n3,dave,23,Paris,hr,91000\n4,ivan,24,Paris,hr,52000\n5,bob,55,Berlin,hr,70000\n6,nora,23,Madrid,hr,126000\n7,grace,56,Tokyo,finance,124000\n8,carol,33,Paris,finance,51000\n9,frank,21,London,finance,132000\n10,karl,39,Paris,eng,58000\n11,alice,64,Berlin,finance,106000\n12,pete,51,Cairo,eng,109000\n\n=== bonus_by_dept ===\ndept,bonus\neng,1000\nfinance,5000\nhr,7000\nops,7000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1180000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0052", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,49,Lisbon,sales,87000\n2,bob,21,Paris,sales,128000\n3,dave,31,Tokyo,ops,80000\n4,mike,48,London,sales,113000\n5,nora,36,Cairo,eng,47000\n6,carol,52,London,marketing,57000\n7,pete,39,Rome,sales,137000\n8,olive,35,London,ops,129000\n9,alice,50,Cairo,sales,124000\n10,heidi,40,Madrid,hr,133000\n11,frank,30,Rome,marketing,122000\n12,eve,49,Tokyo,finance,145000\n13,karl,55,Lisbon,sales,87000\n14,grace,33,Tokyo,hr,99000\n15,judy,64,Tokyo,sales,109000\n16,ivan,42,Tokyo,ops,145000\n17,linda2,31,Berlin,ops,47000\n18,bob2,40,Lisbon,ops,118000\n```\n\nTask: Group the rows by 'city' and compute the max of 'salary' per group. Return the group name with the largest max.", "reference_answer": "Tokyo", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0053", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,162\n1,183\n2,164\n3,165\n4,166\n5,167\n6,168\n7,169\n8,170\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "1", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0054", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,49,Madrid,marketing,50000\n2,bob,63,Madrid,ops,40000\n3,eve,27,Cairo,finance,95000\n4,alice,60,Paris,sales,94000\n5,pete,46,Paris,eng,128000\n6,carol,39,Paris,finance,120000\n7,karl,52,Lisbon,eng,63000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 0 (0-indexed, header excluded)? Return only the value.", "reference_answer": "marketing", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0055", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,heidi,56,Madrid,marketing,144000\n2,carol,38,Lisbon,finance,48000\n3,eve,31,Madrid,sales,75000\n4,olive,40,Cairo,ops,137000\n5,judy,30,Cairo,eng,121000\n6,linda,22,Berlin,eng,54000\n7,pete,43,London,hr,86000\n8,dave,22,Paris,finance,123000\n9,ivan,43,Paris,finance,107000\n10,mike,63,Madrid,hr,123000\n\n=== bonus_by_dept ===\ndept,bonus\neng,4000\nfinance,4000\nhr,5000\nmarketing,6000\nops,7000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1062000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0056", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,60,London,ops,96000\n2,judy,28,Paris,marketing,143000\n3,linda,39,London,eng,95000\n4,alice,33,London,marketing,93000\n5,dave,36,Rome,eng,58000\n6,grace,29,Madrid,finance,100000\n7,frank,45,London,finance,104000\n8,heidi,34,Berlin,finance,81000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 0 (0-indexed, header excluded)? Return only the value.", "reference_answer": "London", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0057", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,62,Lisbon,eng,132000\n2,pete,30,London,sales,101000\n3,grace,28,Rome,sales,112000\n4,judy,56,Paris,hr,102000\n5,mike,23,Cairo,sales,135000\n6,eve,53,Rome,ops,67000\n7,nora,20,London,eng,79000\n8,carol,60,Tokyo,sales,94000\n9,karl,41,Rome,finance,40000\n10,alice,39,Cairo,eng,87000\n11,bob,21,Rome,sales,85000\n12,frank,50,Berlin,ops,47000\n13,olive,42,Rome,marketing,124000\n14,dave,35,Berlin,finance,45000\n```\n\nTask: Compute the mean of the 'age' column. Return the numeric value (any precision).", "reference_answer": "40.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0058", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,25,Berlin,sales,136000\n2,dave,53,Madrid,eng,116000\n3,nora,43,Paris,sales,142000\n4,judy,48,Tokyo,hr,52000\n5,ivan,37,Cairo,sales,77000\n6,linda,42,Paris,marketing,139000\n7,grace,54,Lisbon,finance,92000\n8,frank,46,Berlin,finance,102000\n9,bob,29,London,sales,122000\n10,olive,50,Berlin,sales,123000\n11,alice,46,Cairo,sales,86000\n12,mike,42,Paris,hr,87000\n13,heidi,27,Tokyo,finance,134000\n14,carol,44,Paris,finance,80000\n```\n\nTask: How many rows have age > 42 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0059", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,57\n1,58\n2,59\n3,60\n4,61\n5,62\n6,63\n7,64\n8,65\n9,46\n10,67\n11,68\n12,69\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "9", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0060", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,64,London,finance,127000\n2,frank,29,Rome,ops,41000\n3,dave,46,Paris,hr,107000\n4,grace,61,London,ops,97000\n5,eve,53,Rome,marketing,72000\n6,ivan,22,Rome,finance,64000\n7,heidi,60,Lisbon,ops,97000\n8,carol,20,Berlin,ops,117000\n9,karl,31,London,hr,68000\n10,mike,60,Lisbon,marketing,99000\n11,alice,58,Cairo,finance,45000\n12,olive,20,Tokyo,sales,48000\n13,linda,42,Lisbon,eng,65000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "566", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0061", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,40,Paris,sales,57000\n2,karl,34,Tokyo,ops,130000\n3,heidi,40,Cairo,ops,63000\n4,ivan,61,Madrid,hr,63000\n5,alice,44,Cairo,marketing,120000\n6,bob,51,Berlin,sales,76000\n7,linda,27,London,sales,123000\n8,dave,25,Berlin,finance,57000\n9,frank,29,Cairo,sales,92000\n10,carol,44,Paris,sales,47000\n11,olive,45,Lisbon,marketing,44000\n12,nora,60,Lisbon,finance,125000\n```\n\nTask: How many rows have age > 31 AND city = 'Tokyo'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0062", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,40,Madrid,ops,89000\n2,frank,64,Madrid,hr,44000\n3,alice,60,Tokyo,sales,53000\n4,dave,37,London,finance,59000\n5,judy,44,Lisbon,finance,40000\n6,linda,29,Rome,sales,135000\n7,heidi,33,Cairo,marketing,136000\n8,grace,39,Cairo,sales,140000\n9,ivan,53,Berlin,marketing,124000\n10,carol,44,Berlin,finance,66000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "443", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0063", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,64,Rome,finance,81000\n2,linda,54,Madrid,sales,144000\n3,olive,41,Lisbon,finance,134000\n4,mike,47,Tokyo,finance,74000\n5,eve,26,London,finance,46000\n6,frank,46,Tokyo,finance,73000\n7,judy,29,Berlin,hr,68000\n8,ivan,60,Paris,hr,111000\n9,bob,42,London,sales,104000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,7000\nhr,3000\nsales,5000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "886000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0064", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,173\n1,181\n2,189\n3,197\n4,365\n5,213\n6,221\n7,229\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "4", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0065", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,47,Cairo,ops,109000\n2,alice,36,Paris,eng,51000\n3,nora,26,London,marketing,137000\n4,pete,46,Cairo,finance,57000\n5,judy,28,Madrid,eng,133000\n6,linda,27,Berlin,marketing,131000\n7,grace,47,Lisbon,finance,92000\n8,eve,51,London,finance,117000\n9,dave,38,Madrid,eng,73000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 1 (0-indexed, header excluded)? Return only the value.", "reference_answer": "eng", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0066", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,157\n1,163\n2,169\n3,175\n4,181\n5,187\n6,313\n7,199\n8,205\n9,211\n10,217\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0067", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,58,Paris,sales,95000\n2,dave,39,Madrid,marketing,43000\n3,judy,28,Berlin,finance,81000\n4,eve,31,Lisbon,finance,138000\n5,mike,29,Tokyo,sales,53000\n6,heidi,62,Lisbon,marketing,47000\n7,olive,23,Berlin,hr,46000\n8,carol,63,Tokyo,sales,54000\n9,bob,57,London,eng,75000\n10,pete,48,Madrid,marketing,123000\n11,karl,36,Berlin,hr,60000\n12,ivan,28,Cairo,ops,149000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "37.5", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0068", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,heidi,34,Cairo,marketing,85000\n2,carol,41,Madrid,marketing,60000\n3,nora,61,Rome,hr,59000\n4,judy,43,Madrid,hr,101000\n5,dave,33,Tokyo,hr,120000\n6,karl,41,London,finance,41000\n7,ivan,52,London,hr,115000\n8,linda,42,Lisbon,marketing,74000\n9,mike,40,Berlin,sales,52000\n10,eve,22,Berlin,ops,61000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,5000\nhr,6000\nmarketing,3000\nops,5000\nsales,2000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "813000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0069", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,29,Berlin,sales,142000\n2,carol,63,London,ops,115000\n3,mike,34,Berlin,finance,96000\n4,bob,21,Tokyo,ops,89000\n5,alice,29,Tokyo,hr,44000\n6,grace,41,Cairo,sales,136000\n7,eve,60,London,hr,59000\n8,pete,51,Lisbon,sales,122000\n9,nora,42,Tokyo,eng,127000\n10,olive,53,Rome,ops,66000\n11,ivan,27,Lisbon,hr,99000\n12,frank,20,Berlin,sales,70000\n13,heidi,30,Tokyo,marketing,59000\n14,linda,57,Tokyo,hr,120000\n15,judy,42,Cairo,eng,139000\n```\n\nTask: How many rows have age > 27 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0070", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,47,Madrid,eng,60000\n2,grace,51,Cairo,hr,138000\n3,ivan,62,Lisbon,sales,62000\n4,frank,29,Lisbon,sales,66000\n5,pete,28,Madrid,marketing,97000\n6,nora,55,London,marketing,113000\n7,alice,53,Berlin,marketing,56000\n8,linda,47,Tokyo,marketing,122000\n9,bob,58,Lisbon,marketing,142000\n10,mike,29,London,hr,67000\n11,judy,24,Cairo,sales,68000\n12,dave,36,Lisbon,eng,88000\n13,eve,50,Lisbon,eng,109000\n14,olive,58,Cairo,marketing,124000\n15,carol,37,Madrid,finance,138000\n16,heidi,37,Madrid,sales,44000\n```\n\nTask: How many rows have age > 44 AND city = 'Cairo'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0071", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,56,Tokyo,finance,121000\n2,frank,23,Rome,marketing,117000\n3,heidi,35,Berlin,finance,107000\n4,pete,53,Berlin,marketing,50000\n5,nora,22,Rome,hr,48000\n6,bob,56,Cairo,hr,54000\n7,grace,22,London,ops,43000\n8,olive,60,Lisbon,finance,104000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "327", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0072", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,39,Madrid,eng,139000\n2,pete,38,Cairo,finance,42000\n3,grace,45,Cairo,ops,70000\n4,linda,49,London,sales,137000\n5,bob,24,Rome,marketing,146000\n6,frank,55,Berlin,hr,143000\n7,mike,51,Paris,eng,66000\n8,carol,28,Paris,eng,142000\n9,karl,31,Tokyo,sales,49000\n10,nora,31,Berlin,finance,103000\n11,alice,39,Berlin,ops,97000\n12,judy,45,Lisbon,sales,107000\n13,dave,56,Tokyo,ops,63000\n14,olive,60,Lisbon,eng,97000\n15,heidi,50,Tokyo,eng,50000\n16,ivan,50,Paris,finance,129000\n17,eve2,41,Cairo,hr,142000\n18,pete2,46,Rome,eng,81000\n19,grace2,64,Paris,marketing,141000\n20,linda2,44,London,hr,135000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "Madrid", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0073", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,63,Paris,hr,40000\n2,karl,54,Paris,eng,94000\n3,judy,49,Madrid,sales,103000\n4,eve,45,Madrid,eng,107000\n5,heidi,22,Berlin,marketing,103000\n6,olive,45,London,marketing,92000\n7,nora,49,Rome,hr,113000\n8,mike,36,Madrid,hr,76000\n9,linda,34,Rome,eng,69000\n10,dave,52,Lisbon,ops,47000\n11,bob,59,London,finance,103000\n12,carol,20,Berlin,ops,136000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "136000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0074", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,26,Tokyo,finance,62000\n2,pete,36,Paris,ops,72000\n3,nora,63,Berlin,hr,147000\n4,carol,20,Rome,ops,69000\n5,dave,63,Tokyo,sales,65000\n6,mike,51,Madrid,sales,42000\n7,heidi,38,Cairo,finance,110000\n8,karl,63,Tokyo,hr,74000\n9,alice,31,London,sales,52000\n10,olive,53,Cairo,marketing,117000\n11,frank,35,London,eng,143000\n12,grace,27,Berlin,eng,114000\n```\n\nTask: How many rows have age > 43 AND city = 'Tokyo'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0075", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,20,Rome,marketing,55000\n2,karl,37,Berlin,finance,94000\n3,bob,57,Paris,sales,142000\n4,grace,35,Lisbon,finance,71000\n5,nora,22,Rome,ops,87000\n6,heidi,22,Rome,ops,63000\n7,alice,23,Cairo,sales,56000\n8,pete,28,Rome,marketing,47000\n9,dave,45,London,eng,109000\n10,ivan,52,Berlin,hr,128000\n11,eve,29,Madrid,ops,142000\n12,judy,57,Cairo,marketing,133000\n13,carol,30,Madrid,finance,92000\n14,mike,56,Berlin,finance,144000\n15,olive,60,Lisbon,finance,61000\n16,frank,46,Paris,sales,140000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "hr", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0076", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,62,Tokyo,sales,99000\n2,pete,26,London,hr,47000\n3,olive,61,Paris,hr,91000\n4,alice,42,Cairo,ops,142000\n5,judy,32,Cairo,finance,82000\n6,karl,57,Paris,marketing,75000\n7,heidi,27,Berlin,marketing,71000\n8,bob,25,Tokyo,sales,62000\n9,mike,24,Paris,marketing,54000\n10,eve,32,Rome,eng,128000\n```\n\nTask: Compute the mean of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "85100.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0077", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,64,Madrid,ops,109000\n2,ivan,48,Tokyo,hr,41000\n3,bob,64,Madrid,marketing,89000\n4,mike,58,Rome,sales,99000\n5,eve,37,Rome,hr,58000\n6,dave,37,Cairo,marketing,115000\n7,nora,52,London,marketing,50000\n8,karl,21,London,eng,74000\n9,alice,52,Cairo,finance,78000\n10,heidi,26,Cairo,sales,128000\n11,carol,35,Cairo,marketing,103000\n```\n\nTask: Compute the median of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "89000.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0078", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,ivan,42,Lisbon,hr,95000\n2,nora,49,Tokyo,eng,43000\n3,dave,47,London,sales,75000\n4,pete,57,Lisbon,eng,56000\n5,eve,44,Paris,hr,84000\n6,grace,64,Madrid,hr,145000\n7,frank,62,Tokyo,ops,44000\n8,heidi,44,London,eng,104000\n9,karl,38,Paris,marketing,75000\n10,alice,26,London,hr,74000\n11,bob,51,Lisbon,marketing,144000\n\n=== bonus_by_dept ===\ndept,bonus\neng,6000\nhr,9000\nmarketing,4000\nops,1000\nsales,6000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1008000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0079", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,133\n1,301\n2,149\n3,157\n4,165\n5,173\n6,181\n7,189\n8,197\n9,205\n10,213\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "1", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0080", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,29,Cairo,sales,60000\n2,karl,34,Cairo,ops,40000\n3,heidi,44,Tokyo,eng,55000\n4,olive,50,Madrid,eng,123000\n5,eve,24,Berlin,hr,52000\n6,grace,25,Cairo,eng,72000\n7,bob,25,Cairo,finance,89000\n8,linda,54,Rome,eng,81000\n9,nora,25,Rome,finance,147000\n10,frank,50,Paris,marketing,44000\n11,pete,47,London,sales,71000\n12,alice,37,Cairo,hr,134000\n13,ivan,61,Berlin,hr,47000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "marketing", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0081", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,175\n1,182\n2,189\n3,196\n4,203\n5,210\n6,217\n7,364\n8,231\n9,238\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "7", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0082", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,30,London,sales,81000\n2,dave,27,Lisbon,sales,131000\n3,linda,44,London,eng,71000\n4,frank,63,Berlin,ops,133000\n5,karl,53,Berlin,finance,131000\n6,nora,52,Tokyo,marketing,143000\n7,alice,56,Madrid,eng,61000\n8,ivan,64,Lisbon,sales,40000\n9,pete,27,Lisbon,ops,50000\n10,heidi,32,Madrid,eng,79000\n11,carol,20,Paris,marketing,46000\n12,olive,43,Paris,eng,89000\n13,bob,24,Lisbon,hr,64000\n14,eve,22,Paris,marketing,44000\n15,judy,50,Madrid,finance,111000\n16,grace,27,Berlin,finance,65000\n17,mike2,51,Tokyo,hr,139000\n18,dave2,57,Paris,sales,123000\n19,linda2,27,Tokyo,finance,56000\n```\n\nTask: Group the rows by 'dept' and compute the max of 'age' per group. Return the group name with the largest max.", "reference_answer": "sales", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0083", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,26,Paris,marketing,135000\n2,carol,62,Paris,sales,77000\n3,grace,26,Berlin,hr,71000\n4,judy,58,Tokyo,finance,83000\n5,linda,41,Lisbon,finance,44000\n6,bob,33,Cairo,ops,136000\n7,alice,46,Cairo,hr,111000\n8,ivan,39,Lisbon,hr,56000\n9,heidi,45,London,marketing,115000\n10,mike,22,Madrid,hr,68000\n11,dave,32,London,hr,73000\n12,eve,53,Tokyo,finance,123000\n13,karl,35,Lisbon,marketing,57000\n14,pete,37,Rome,eng,75000\n15,olive,42,Madrid,ops,144000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "Tokyo", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0084", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,54,Cairo,ops,61000\n2,karl,30,Tokyo,ops,136000\n3,linda,41,Tokyo,hr,102000\n4,bob,61,Madrid,finance,113000\n5,alice,59,Tokyo,hr,107000\n6,olive,28,Paris,ops,146000\n7,grace,45,Madrid,ops,102000\n8,mike,55,Lisbon,hr,140000\n9,nora,41,Berlin,eng,130000\n10,dave,58,Madrid,finance,94000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 8 (0-indexed, header excluded)? Return only the value.", "reference_answer": "eng", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0085", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,40,Madrid,eng,111000\n2,ivan,42,Lisbon,ops,148000\n3,dave,63,Cairo,ops,134000\n4,karl,42,Berlin,ops,92000\n5,linda,61,Lisbon,sales,129000\n6,alice,30,Lisbon,eng,137000\n7,carol,56,Paris,sales,142000\n8,nora,59,Cairo,eng,85000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 2 (0-indexed, header excluded)? Return only the value.", "reference_answer": "63", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0086", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,35,London,marketing,104000\n2,mike,49,Tokyo,finance,109000\n3,karl,41,Rome,sales,102000\n4,dave,47,Rome,ops,82000\n5,frank,58,London,marketing,80000\n6,judy,41,Rome,hr,61000\n7,olive,31,Rome,sales,121000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 2 (0-indexed, header excluded)? Return only the value.", "reference_answer": "sales", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0087", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,42,Berlin,marketing,131000\n2,alice,35,Paris,eng,149000\n3,frank,54,Lisbon,hr,82000\n4,karl,28,Madrid,ops,87000\n5,bob,40,Lisbon,sales,147000\n6,mike,23,Madrid,finance,110000\n7,olive,33,Cairo,eng,118000\n8,grace,42,London,marketing,71000\n9,heidi,25,Rome,hr,77000\n10,pete,24,Paris,finance,149000\n11,eve,49,Cairo,ops,149000\n12,ivan,38,Cairo,hr,51000\n13,nora,59,Rome,eng,125000\n14,judy,45,Berlin,eng,96000\n15,carol,57,Cairo,finance,102000\n16,linda,61,Lisbon,eng,121000\n17,dave2,47,Lisbon,ops,45000\n18,alice2,46,Cairo,sales,71000\n19,frank2,43,Rome,sales,110000\n20,karl2,23,Paris,ops,106000\n21,bob2,27,Berlin,eng,125000\n22,mike2,28,Lisbon,eng,92000\n```\n\nTask: How many rows have age > 48 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0088", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,bob,45,Madrid,ops,107000\n2,alice,42,Berlin,sales,88000\n3,frank,56,Paris,sales,60000\n4,eve,22,Cairo,finance,140000\n5,carol,43,Cairo,hr,59000\n6,ivan,32,London,ops,96000\n7,nora,30,Berlin,finance,89000\n8,grace,46,Cairo,sales,147000\n9,heidi,33,Madrid,marketing,91000\n10,linda,31,Rome,eng,51000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "928000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0089", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,30,Paris,finance,117000\n2,pete,52,Paris,marketing,114000\n3,olive,61,Lisbon,marketing,78000\n4,nora,47,Madrid,finance,71000\n5,heidi,20,Berlin,marketing,62000\n6,karl,32,Madrid,marketing,146000\n7,eve,36,Lisbon,finance,142000\n8,linda,61,Lisbon,eng,136000\n9,judy,62,Paris,ops,142000\n10,carol,49,London,eng,91000\n11,alice,36,Cairo,finance,52000\n12,bob,32,Lisbon,marketing,125000\n13,frank,34,Madrid,ops,138000\n14,mike,41,Cairo,eng,98000\n15,grace,45,Rome,hr,92000\n16,ivan,34,London,finance,53000\n17,dave2,32,Paris,ops,91000\n18,pete2,47,Tokyo,hr,130000\n19,olive2,63,Lisbon,ops,46000\n20,nora2,56,Tokyo,eng,118000\n21,heidi2,30,London,marketing,118000\n22,karl2,53,London,eng,66000\n23,eve2,55,Berlin,marketing,115000\n24,linda2,24,Lisbon,marketing,121000\n```\n\nTask: How many rows have age > 37 AND city = 'London'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0090", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,48,London,eng,116000\n2,judy,28,Lisbon,finance,58000\n3,frank,57,Tokyo,marketing,60000\n4,grace,33,Rome,ops,92000\n5,bob,35,Tokyo,sales,141000\n6,olive,54,Cairo,marketing,100000\n7,heidi,27,Cairo,hr,88000\n8,dave,64,Lisbon,sales,140000\n9,karl,40,Berlin,hr,141000\n10,nora,47,Tokyo,finance,40000\n```\n\nTask: In the CSV below, what is the value of column 'salary' in row 1 (0-indexed, header excluded)? Return only the value.", "reference_answer": "58000", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0091", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,56,Lisbon,sales,43000\n2,frank,34,Madrid,sales,103000\n3,ivan,62,Paris,marketing,66000\n4,eve,35,Madrid,ops,123000\n5,bob,37,Tokyo,finance,147000\n6,grace,56,Madrid,finance,105000\n7,linda,28,Rome,finance,127000\n8,judy,48,London,eng,80000\n9,mike,27,Paris,hr,125000\n10,heidi,39,Rome,hr,80000\n11,karl,44,Lisbon,marketing,65000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "147000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0092", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,36,Berlin,hr,132000\n2,pete,58,Berlin,eng,46000\n3,carol,31,Tokyo,ops,40000\n4,mike,40,Lisbon,finance,55000\n5,alice,21,Madrid,sales,108000\n6,judy,28,Cairo,ops,53000\n7,ivan,47,London,ops,136000\n8,frank,41,Madrid,ops,42000\n9,eve,24,Lisbon,marketing,149000\n10,nora,31,Cairo,marketing,80000\n11,linda,30,Berlin,eng,112000\n12,olive,37,Tokyo,eng,58000\n13,dave,52,Cairo,hr,102000\n14,bob,20,Berlin,sales,143000\n15,grace,55,Lisbon,ops,69000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "149000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0093", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,25,Berlin,finance,76000\n2,olive,40,Paris,ops,61000\n3,ivan,22,Berlin,hr,105000\n4,judy,28,Tokyo,marketing,65000\n5,grace,47,Lisbon,hr,67000\n6,dave,41,Berlin,finance,129000\n7,pete,21,Cairo,ops,43000\n8,bob,37,Lisbon,ops,130000\n9,nora,64,Tokyo,finance,86000\n10,mike,45,Rome,hr,128000\n```\n\nTask: In the CSV below, what is the value of column 'salary' in row 0 (0-indexed, header excluded)? Return only the value.", "reference_answer": "76000", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0094", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,41,London,ops,143000\n2,heidi,45,Rome,sales,64000\n3,linda,34,London,eng,121000\n4,alice,24,London,hr,126000\n5,carol,41,Lisbon,finance,61000\n6,judy,62,Cairo,eng,100000\n7,eve,26,Berlin,hr,71000\n8,ivan,60,Rome,finance,68000\n9,dave,34,Madrid,eng,100000\n10,bob,23,Berlin,hr,132000\n11,frank,53,Tokyo,hr,114000\n12,grace,46,Madrid,marketing,51000\n13,karl,31,Berlin,finance,72000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "Cairo", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0095", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,46,Cairo,ops,62000\n2,heidi,28,Berlin,finance,71000\n3,mike,35,London,finance,69000\n4,alice,27,London,marketing,69000\n5,ivan,31,Cairo,hr,130000\n6,bob,25,Cairo,eng,53000\n7,nora,61,Rome,finance,49000\n8,pete,32,Lisbon,hr,137000\n9,frank,28,Cairo,eng,98000\n10,karl,28,Cairo,hr,43000\n11,dave,26,Cairo,hr,122000\n12,linda,61,Madrid,marketing,134000\n13,grace,24,Cairo,finance,91000\n14,olive,40,Madrid,eng,45000\n15,eve,59,Lisbon,eng,70000\n16,carol,62,Tokyo,marketing,145000\n17,judy2,29,Madrid,ops,71000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "31.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0096", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,40,Paris,marketing,85000\n2,grace,28,Cairo,marketing,46000\n3,mike,24,Lisbon,hr,40000\n4,karl,40,Paris,sales,122000\n5,olive,37,Tokyo,finance,78000\n6,ivan,44,Lisbon,marketing,89000\n7,dave,41,Cairo,hr,82000\n8,heidi,34,Rome,hr,99000\n9,nora,57,Madrid,finance,87000\n10,frank,29,Rome,marketing,50000\n11,bob,45,Tokyo,eng,46000\n12,carol,54,Rome,marketing,47000\n13,alice,55,Berlin,sales,96000\n14,linda,47,Tokyo,hr,90000\n15,pete,33,Rome,eng,104000\n```\n\nTask: How many rows have age > 43 AND city = 'Rome'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0097", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,62,Madrid,marketing,122000\n2,mike,53,Cairo,marketing,107000\n3,pete,30,Cairo,eng,76000\n4,frank,21,Tokyo,eng,121000\n5,grace,21,Paris,ops,102000\n6,ivan,36,Cairo,hr,107000\n7,heidi,50,Madrid,hr,145000\n8,bob,28,Tokyo,sales,122000\n9,judy,33,Madrid,hr,59000\n10,dave,26,Rome,sales,100000\n11,olive,64,Rome,sales,46000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "424", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0098", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,157\n1,162\n2,167\n3,172\n4,177\n5,182\n6,87\n7,192\n8,197\n9,202\n10,207\n11,212\n12,217\n13,222\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0099", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,karl,20,Tokyo,ops,56000\n2,linda,36,Rome,ops,65000\n3,frank,41,Cairo,hr,57000\n4,judy,39,Berlin,hr,119000\n5,pete,46,Lisbon,eng,55000\n6,dave,28,Berlin,finance,115000\n7,nora,24,Cairo,sales,126000\n8,ivan,59,Cairo,sales,57000\n9,carol,29,Madrid,eng,102000\n10,bob,51,Berlin,ops,146000\n11,mike,43,Cairo,marketing,117000\n12,grace,23,Berlin,finance,58000\n13,heidi,45,Paris,finance,65000\n\n=== bonus_by_dept ===\ndept,bonus\neng,8000\nfinance,5000\nhr,9000\nmarketing,4000\nops,8000\nsales,7000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1229000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0100", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,63\n1,65\n2,67\n3,69\n4,111\n5,73\n6,75\n7,77\n8,79\n9,81\n10,83\n11,85\n12,87\n13,89\n14,91\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "4", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0101", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,23,Lisbon,marketing,48000\n2,heidi,63,Lisbon,hr,121000\n3,frank,43,Madrid,eng,124000\n4,carol,49,London,finance,119000\n5,olive,53,Tokyo,hr,52000\n6,bob,50,Cairo,sales,43000\n7,judy,20,Tokyo,sales,129000\n8,alice,61,Madrid,finance,44000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 2 (0-indexed, header excluded)? Return only the value.", "reference_answer": "eng", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0102", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,51,Cairo,sales,125000\n2,pete,23,Paris,eng,87000\n3,alice,46,London,ops,67000\n4,ivan,29,Berlin,eng,43000\n5,bob,35,Tokyo,eng,127000\n6,linda,61,Cairo,marketing,127000\n7,mike,60,Madrid,finance,108000\n8,nora,52,Paris,hr,111000\n9,heidi,53,Rome,hr,131000\n10,eve,21,Lisbon,ops,83000\n11,dave,49,Lisbon,finance,79000\n12,karl,47,Cairo,eng,94000\n13,frank,29,Lisbon,ops,83000\n14,olive,63,Lisbon,ops,76000\n15,carol,53,Cairo,hr,121000\n16,judy,45,Lisbon,hr,103000\n17,grace2,42,Tokyo,marketing,142000\n18,pete2,37,London,finance,106000\n19,alice2,43,Madrid,ops,141000\n20,ivan2,33,Tokyo,finance,62000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "Rome", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0103", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,56,Tokyo,hr,54000\n2,karl,31,Tokyo,sales,141000\n3,dave,31,Berlin,finance,112000\n4,grace,51,Rome,hr,85000\n5,bob,41,London,finance,142000\n6,judy,64,Tokyo,marketing,145000\n7,alice,29,Rome,finance,142000\n8,olive,30,Berlin,hr,144000\n9,mike,58,Paris,finance,54000\n10,linda,39,Tokyo,eng,113000\n11,heidi,56,London,sales,101000\n12,nora,32,Paris,marketing,63000\n13,frank,61,Paris,eng,135000\n14,pete,20,Tokyo,eng,47000\n15,eve,62,Rome,ops,109000\n```\n\nTask: How many rows have age > 46 AND city = 'Rome'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0104", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,bob,58,Tokyo,ops,130000\n2,alice,33,Madrid,eng,72000\n3,mike,23,Rome,ops,87000\n4,pete,55,Paris,sales,135000\n5,judy,58,London,marketing,70000\n6,carol,42,Paris,sales,102000\n7,ivan,47,Tokyo,marketing,43000\n8,grace,54,Madrid,finance,120000\n9,olive,32,Madrid,finance,127000\n10,karl,28,Paris,marketing,63000\n11,linda,54,Cairo,ops,59000\n12,eve,45,Rome,hr,125000\n13,heidi,27,Lisbon,hr,84000\n14,frank,27,London,ops,119000\n15,nora,57,Tokyo,eng,57000\n16,dave,20,London,ops,110000\n```\n\nTask: How many rows have age > 42 AND city = 'Madrid'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0105", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,60,Lisbon,ops,72000\n2,olive,56,Madrid,finance,104000\n3,grace,58,London,sales,126000\n4,linda,25,Tokyo,hr,51000\n5,carol,60,Lisbon,sales,100000\n6,bob,60,Tokyo,marketing,50000\n7,karl,43,Cairo,eng,143000\n8,alice,53,Paris,sales,50000\n9,pete,57,Paris,hr,111000\n10,frank,47,Paris,finance,72000\n11,eve,23,Tokyo,sales,49000\n12,dave,25,Madrid,eng,96000\n13,nora,57,Rome,eng,110000\n14,heidi,21,Tokyo,ops,78000\n15,judy,25,Tokyo,hr,137000\n16,mike,57,Paris,eng,141000\n17,ivan2,54,Paris,sales,113000\n18,olive2,25,Madrid,eng,76000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "marketing", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0106", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,35,Paris,hr,71000\n2,nora,23,Tokyo,hr,123000\n3,frank,40,Berlin,marketing,111000\n4,carol,58,Cairo,hr,49000\n5,eve,21,Madrid,eng,76000\n6,linda,48,Lisbon,sales,92000\n7,dave,23,Paris,eng,113000\n8,ivan,61,London,finance,46000\n9,karl,59,Cairo,finance,89000\n10,olive,27,London,hr,115000\n11,bob,36,Cairo,ops,45000\n12,mike,31,Cairo,sales,64000\n```\n\nTask: Group the rows by 'dept' and compute the max of 'age' per group. Return the group name with the largest max.", "reference_answer": "finance", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0107", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,25,Rome,ops,82000\n2,eve,24,Madrid,hr,112000\n3,ivan,49,Berlin,finance,46000\n4,bob,39,Madrid,finance,146000\n5,dave,56,Tokyo,ops,102000\n6,alice,24,Cairo,sales,123000\n7,olive,22,Cairo,marketing,70000\n8,grace,63,London,sales,59000\n9,pete,46,Lisbon,marketing,95000\n10,judy,26,Rome,ops,78000\n11,heidi,62,London,ops,73000\n12,frank,28,Cairo,hr,64000\n13,linda,24,Cairo,eng,120000\n14,carol,48,Rome,eng,105000\n15,karl,34,Madrid,finance,53000\n16,nora,49,Cairo,ops,105000\n17,mike2,62,Berlin,ops,91000\n18,eve2,29,London,hr,63000\n19,ivan2,39,Cairo,hr,74000\n20,bob2,59,Paris,marketing,73000\n21,dave2,36,Paris,marketing,44000\n22,alice2,61,Berlin,hr,77000\n```\n\nTask: How many rows have age > 31 AND city = 'Paris'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0108", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,22,Berlin,ops,147000\n2,pete,22,London,sales,59000\n3,linda,49,Madrid,marketing,132000\n4,alice,52,Madrid,marketing,75000\n5,olive,56,Lisbon,eng,96000\n6,grace,29,London,eng,63000\n7,carol,28,Tokyo,finance,68000\n8,karl,33,Berlin,hr,75000\n9,frank,56,Rome,finance,123000\n10,ivan,21,Berlin,ops,69000\n11,heidi,55,Berlin,eng,127000\n12,eve,38,London,ops,130000\n13,judy,54,Lisbon,ops,137000\n14,bob,51,Berlin,finance,88000\n15,mike,27,London,eng,71000\n16,dave,37,Tokyo,sales,52000\n17,nora2,25,Paris,sales,101000\n18,pete2,29,Madrid,finance,133000\n19,linda2,31,Berlin,eng,59000\n20,alice2,56,Cairo,eng,57000\n21,olive2,49,Berlin,eng,79000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "ops", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0109", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,28,Berlin,marketing,76000\n2,eve,37,Paris,marketing,41000\n3,alice,56,Tokyo,eng,114000\n4,karl,39,Rome,sales,145000\n5,nora,39,Paris,sales,96000\n6,dave,56,Cairo,sales,109000\n7,linda,59,Rome,ops,93000\n8,frank,49,London,hr,149000\n9,judy,61,Rome,marketing,43000\n10,grace,53,Paris,sales,102000\n11,olive,46,Lisbon,sales,65000\n12,heidi,46,Rome,marketing,71000\n13,ivan,57,Rome,eng,112000\n14,pete,25,Paris,eng,84000\n15,carol,53,Lisbon,sales,78000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1378000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0110", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,dave,52,Tokyo,sales,64000\n2,pete,61,London,hr,44000\n3,linda,45,Tokyo,finance,108000\n4,bob,21,Rome,marketing,148000\n5,ivan,22,Berlin,sales,63000\n6,olive,64,Lisbon,sales,136000\n7,carol,54,Tokyo,sales,141000\n8,heidi,46,Rome,ops,104000\n9,grace,50,London,hr,148000\n10,mike,32,Cairo,hr,58000\n11,eve,46,Madrid,finance,148000\n12,karl,50,Tokyo,sales,142000\n13,alice,32,Berlin,sales,102000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,4000\nhr,4000\nmarketing,5000\nops,6000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1485000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0111", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,59,Lisbon,hr,100000\n2,pete,62,Lisbon,finance,107000\n3,bob,34,London,finance,75000\n4,alice,48,Tokyo,ops,136000\n5,frank,42,Cairo,hr,116000\n6,carol,50,London,marketing,56000\n7,ivan,29,Tokyo,hr,147000\n8,heidi,64,Rome,marketing,119000\n9,mike,26,London,eng,94000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 3 (0-indexed, header excluded)? Return only the value.", "reference_answer": "48", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0112", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,30,Cairo,finance,98000\n2,heidi,45,Berlin,eng,133000\n3,linda,45,Lisbon,finance,140000\n4,nora,40,Berlin,marketing,120000\n5,alice,43,London,hr,107000\n6,frank,23,Madrid,sales,112000\n7,grace,20,Tokyo,finance,44000\n8,olive,58,Berlin,hr,60000\n9,eve,28,Cairo,hr,100000\n10,carol,44,Madrid,eng,84000\n11,ivan,53,Lisbon,marketing,105000\n12,mike,31,Berlin,finance,134000\n13,dave,30,Cairo,finance,82000\n14,judy,21,Madrid,ops,47000\n15,bob,52,Rome,marketing,96000\n16,pete,43,Lisbon,hr,111000\n17,karl2,28,London,eng,117000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "40.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0113", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,36,Rome,ops,54000\n2,pete,57,Paris,hr,145000\n3,eve,63,Lisbon,finance,112000\n4,olive,56,Madrid,eng,139000\n5,alice,24,Rome,sales,66000\n6,dave,29,Berlin,sales,54000\n7,nora,49,Rome,eng,45000\n8,bob,64,Lisbon,sales,81000\n9,mike,51,Madrid,hr,136000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 5 (0-indexed, header excluded)? Return only the value.", "reference_answer": "Berlin", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0114", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,28,Lisbon,sales,99000\n2,frank,39,Berlin,eng,64000\n3,carol,30,Cairo,sales,57000\n4,karl,54,Rome,marketing,84000\n5,alice,42,Cairo,hr,88000\n6,nora,41,Berlin,ops,146000\n7,grace,40,Tokyo,eng,56000\n8,linda,40,Berlin,finance,44000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "638000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0115", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,61,London,eng,138000\n2,frank,47,Berlin,hr,56000\n3,linda,25,Lisbon,ops,70000\n4,alice,32,Cairo,eng,120000\n5,judy,58,Paris,eng,77000\n6,nora,21,Rome,finance,74000\n7,eve,29,Rome,sales,101000\n8,heidi,29,Cairo,hr,98000\n9,karl,56,Berlin,finance,147000\n10,ivan,56,London,ops,66000\n11,grace,52,Paris,eng,58000\n12,pete,37,Lisbon,ops,49000\n13,olive,53,Paris,eng,146000\n14,mike,39,London,ops,95000\n15,bob,60,Madrid,ops,116000\n```\n\nTask: Compute the mean of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "94066.66666666667", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0116", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,59,Cairo,hr,45000\n2,carol,20,Lisbon,ops,48000\n3,mike,22,Berlin,ops,78000\n4,grace,26,Madrid,eng,130000\n5,pete,42,Lisbon,sales,82000\n6,heidi,26,Paris,ops,100000\n7,nora,44,Paris,marketing,89000\n8,eve,49,Rome,ops,144000\n9,olive,27,Paris,hr,123000\n10,ivan,27,Cairo,hr,111000\n11,dave,39,Lisbon,marketing,49000\n12,karl,52,Berlin,sales,73000\n```\n\nTask: Group the rows by 'city' and compute the max of 'salary' per group. Return the group name with the largest max.", "reference_answer": "Rome", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0117", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,84\n1,91\n2,98\n3,-35\n4,112\n5,119\n6,126\n7,133\n8,140\n9,147\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "3", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0118", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,26,London,eng,123000\n2,carol,26,Berlin,hr,125000\n3,bob,33,Paris,marketing,44000\n4,grace,47,Tokyo,marketing,106000\n5,alice,38,London,marketing,75000\n6,heidi,48,Rome,marketing,128000\n7,dave,30,Tokyo,sales,127000\n8,ivan,36,Lisbon,eng,94000\n9,judy,64,London,finance,114000\n10,linda,32,Berlin,marketing,45000\n11,pete,25,London,eng,82000\n12,eve,20,Tokyo,finance,48000\n13,karl,41,Paris,sales,119000\n14,olive,58,Madrid,marketing,107000\n15,nora,33,Madrid,eng,109000\n16,frank,53,Tokyo,finance,79000\n```\n\nTask: How many rows have age > 41 AND city = 'London'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0119", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,32,Madrid,eng,128000\n2,pete,26,Lisbon,ops,40000\n3,dave,53,London,finance,136000\n4,olive,51,Berlin,sales,110000\n5,alice,32,Cairo,sales,139000\n6,ivan,46,London,ops,75000\n7,eve,35,Cairo,marketing,86000\n8,judy,25,Rome,hr,56000\n9,carol,27,Lisbon,sales,146000\n10,frank,59,Tokyo,hr,59000\n11,heidi,64,Tokyo,marketing,91000\n12,karl,58,Cairo,sales,48000\n```\n\nTask: Compute the mean of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "92833.33333333333", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0120", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,nora,45,Tokyo,ops,134000\n2,grace,48,Cairo,sales,122000\n3,judy,40,Paris,marketing,149000\n4,alice,62,Rome,finance,147000\n5,mike,38,Lisbon,hr,120000\n6,bob,25,Cairo,finance,47000\n7,karl,39,Paris,eng,80000\n8,dave,42,London,eng,131000\n9,ivan,53,Paris,ops,54000\n10,olive,32,Cairo,hr,116000\n11,carol,42,London,sales,147000\n12,eve,48,Rome,hr,146000\n13,frank,36,Lisbon,eng,139000\n14,heidi,53,Rome,ops,93000\n15,linda,28,Berlin,ops,47000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "149000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0121", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,dave,20,Tokyo,eng,149000\n2,ivan,62,Paris,hr,91000\n3,heidi,58,Lisbon,hr,107000\n4,carol,24,Lisbon,eng,63000\n5,alice,21,Lisbon,marketing,101000\n6,frank,52,Rome,marketing,147000\n7,olive,54,Cairo,eng,146000\n8,linda,58,Lisbon,marketing,124000\n9,pete,55,Rome,ops,117000\n10,mike,59,Cairo,eng,55000\n11,eve,50,London,sales,57000\n12,nora,33,Lisbon,ops,147000\n\n=== bonus_by_dept ===\ndept,bonus\neng,9000\nhr,4000\nmarketing,3000\nops,9000\nsales,6000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1381000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0122", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,125\n1,126\n2,127\n3,128\n4,129\n5,150\n6,131\n7,132\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "5", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0123", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,29,Berlin,hr,129000\n2,ivan,56,Rome,eng,109000\n3,olive,56,Rome,eng,133000\n4,judy,20,London,sales,113000\n5,mike,43,Cairo,eng,118000\n6,heidi,52,Madrid,marketing,73000\n7,frank,29,Lisbon,ops,105000\n8,karl,53,Cairo,sales,120000\n9,dave,29,Rome,eng,97000\n```\n\nTask: Compute the median of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "113000.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0124", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,bob,23,Berlin,ops,62000\n2,karl,39,Madrid,ops,81000\n3,eve,39,London,ops,144000\n4,linda,64,Lisbon,ops,120000\n5,judy,55,Madrid,eng,149000\n6,alice,60,Berlin,sales,134000\n7,dave,57,Tokyo,finance,83000\n8,heidi,56,Rome,ops,47000\n9,grace,60,London,marketing,44000\n\n=== bonus_by_dept ===\ndept,bonus\neng,2000\nfinance,5000\nmarketing,7000\nops,1000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "884000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0125", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,linda,31,London,finance,74000\n2,bob,49,Berlin,hr,77000\n3,frank,23,Tokyo,finance,71000\n4,grace,51,Lisbon,marketing,70000\n5,carol,22,London,ops,128000\n6,alice,60,Paris,hr,90000\n7,eve,30,London,sales,61000\n8,nora,25,London,ops,142000\n9,mike,60,Lisbon,sales,128000\n10,karl,60,Paris,ops,130000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,1000\nhr,9000\nmarketing,5000\nops,4000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1024000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0126", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,156\n1,158\n2,160\n3,162\n4,164\n5,166\n6,208\n7,170\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0127", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,57,Rome,marketing,130000\n2,linda,28,Madrid,ops,51000\n3,olive,36,Tokyo,ops,148000\n4,eve,45,London,ops,101000\n5,nora,32,Berlin,ops,121000\n6,bob,28,Lisbon,finance,120000\n7,grace,22,Paris,finance,101000\n8,karl,34,London,marketing,149000\n9,frank,26,Berlin,marketing,127000\n10,carol,53,Tokyo,ops,134000\n11,pete,40,Cairo,hr,66000\n12,ivan,54,London,marketing,142000\n13,heidi,58,Paris,finance,103000\n14,judy,46,Tokyo,hr,139000\n15,mike,56,Tokyo,marketing,86000\n16,alice,59,Paris,eng,85000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "marketing", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0128", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,28,Rome,hr,112000\n2,frank,32,Tokyo,sales,122000\n3,mike,41,Rome,finance,121000\n4,alice,62,Cairo,finance,93000\n5,nora,43,Lisbon,ops,111000\n6,dave,54,London,eng,51000\n7,bob,21,London,hr,41000\n8,heidi,39,London,sales,92000\n9,karl,58,Berlin,finance,109000\n10,pete,50,Paris,eng,104000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "956000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0129", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,bob,30,London,eng,92000\n2,carol,29,Rome,finance,127000\n3,nora,31,London,ops,83000\n4,grace,47,Paris,eng,88000\n5,frank,52,London,marketing,144000\n6,eve,63,Lisbon,hr,88000\n7,alice,39,Paris,finance,82000\n8,linda,33,Lisbon,marketing,66000\n9,olive,47,London,eng,50000\n\n=== bonus_by_dept ===\ndept,bonus\neng,6000\nfinance,3000\nhr,1000\nmarketing,3000\nops,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "859000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0130", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,31,Madrid,sales,146000\n2,carol,27,London,ops,138000\n3,nora,62,Cairo,ops,73000\n4,judy,54,Rome,marketing,113000\n5,pete,47,Madrid,sales,137000\n6,bob,50,Lisbon,hr,89000\n7,alice,64,Lisbon,finance,48000\n8,eve,41,Madrid,sales,82000\n9,heidi,20,Madrid,sales,85000\n10,grace,24,London,sales,146000\n11,ivan,57,London,hr,76000\n12,karl,26,Cairo,marketing,48000\n```\n\nTask: How many rows have age > 28 AND city = 'Madrid'? Return the integer count.", "reference_answer": "3", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0131", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,53,Madrid,hr,52000\n2,nora,56,Madrid,eng,52000\n3,grace,27,Berlin,ops,54000\n4,heidi,24,Rome,hr,135000\n5,mike,49,Tokyo,eng,65000\n6,olive,63,Madrid,marketing,78000\n7,linda,28,Rome,ops,125000\n8,carol,24,Rome,marketing,71000\n9,alice,63,Lisbon,finance,96000\n10,frank,44,Tokyo,eng,69000\n11,ivan,39,Lisbon,hr,109000\n12,pete,41,Lisbon,marketing,58000\n13,dave,32,Cairo,sales,148000\n```\n\nTask: Group the rows by 'city' and compute the max of 'age' per group. Return the group name with the largest max.", "reference_answer": "Madrid", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0132", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,24,London,marketing,81000\n2,alice,26,Berlin,finance,147000\n3,grace,48,Lisbon,hr,146000\n4,linda,31,Lisbon,sales,43000\n5,bob,30,Lisbon,eng,99000\n6,heidi,22,Berlin,ops,141000\n7,olive,24,Paris,ops,108000\n8,eve,38,Lisbon,sales,53000\n9,dave,28,Rome,finance,70000\n10,pete,50,Tokyo,finance,52000\n```\n\nTask: Compute the median of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "90000.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0133", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,bob,60,Cairo,sales,136000\n2,karl,39,Paris,finance,42000\n3,pete,36,Rome,hr,124000\n4,olive,28,Berlin,hr,100000\n5,nora,52,Cairo,eng,47000\n6,ivan,37,Madrid,eng,73000\n7,linda,30,Berlin,marketing,57000\n8,carol,41,Lisbon,ops,104000\n9,frank,31,London,hr,113000\n10,dave,30,Paris,finance,138000\n11,judy,40,London,sales,91000\n\n=== bonus_by_dept ===\ndept,bonus\neng,9000\nfinance,4000\nhr,2000\nmarketing,4000\nops,5000\nsales,5000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1076000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0134", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,20,London,hr,113000\n2,linda,51,Lisbon,hr,147000\n3,ivan,26,Madrid,eng,66000\n4,mike,21,Tokyo,marketing,106000\n5,frank,28,Paris,finance,71000\n6,judy,26,Paris,marketing,52000\n7,bob,21,Paris,hr,114000\n8,pete,29,Tokyo,eng,71000\n9,grace,29,Madrid,eng,101000\n10,eve,43,Cairo,ops,100000\n11,nora,43,Madrid,ops,111000\n12,carol,49,Tokyo,marketing,73000\n13,dave,30,Madrid,hr,66000\n```\n\nTask: Compute the max of the 'age' column. Return the integer.", "reference_answer": "51", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0135", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,45,Rome,ops,43000\n2,carol,41,Rome,finance,143000\n3,dave,64,Cairo,ops,56000\n4,judy,57,Paris,hr,103000\n5,alice,43,Lisbon,ops,89000\n6,bob,46,Berlin,sales,93000\n7,ivan,63,London,ops,140000\n8,pete,42,London,hr,57000\n9,olive,28,Paris,marketing,80000\n10,mike,46,Tokyo,ops,92000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,7000\nhr,7000\nmarketing,7000\nops,4000\nsales,6000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "950000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0136", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,37,Paris,ops,114000\n2,carol,63,Rome,eng,86000\n3,linda,48,Paris,hr,48000\n4,eve,33,Paris,eng,61000\n5,olive,41,Paris,sales,91000\n6,pete,24,Madrid,finance,40000\n7,nora,38,Lisbon,hr,125000\n8,karl,39,Berlin,eng,139000\n9,alice,25,Madrid,marketing,78000\n10,bob,23,London,eng,128000\n11,judy,22,Paris,hr,57000\n12,ivan,25,London,ops,90000\n13,frank,40,Lisbon,finance,149000\n14,mike,47,Lisbon,hr,83000\n15,dave,62,Berlin,sales,110000\n16,heidi,43,Tokyo,marketing,40000\n17,grace2,42,Cairo,sales,42000\n18,carol2,37,Cairo,finance,137000\n19,linda2,43,Cairo,finance,107000\n20,eve2,55,Tokyo,finance,111000\n21,olive2,45,Tokyo,hr,116000\n22,pete2,21,Lisbon,hr,137000\n```\n\nTask: How many rows have age > 40 AND city = 'Madrid'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0137", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,58,Tokyo,finance,145000\n2,bob,25,Paris,finance,99000\n3,linda,51,Cairo,hr,73000\n4,olive,56,Rome,marketing,59000\n5,frank,31,Paris,finance,67000\n6,pete,51,Rome,marketing,121000\n7,heidi,57,Tokyo,marketing,118000\n8,judy,56,Rome,ops,82000\n9,grace,21,Rome,hr,128000\n10,eve,36,Lisbon,finance,51000\n11,nora,34,Lisbon,hr,145000\n12,alice,35,Cairo,ops,62000\n13,carol,61,London,marketing,142000\n14,karl,22,Paris,ops,78000\n15,dave,23,London,marketing,139000\n16,ivan,39,Lisbon,finance,115000\n```\n\nTask: How many rows have age > 49 AND city = 'Paris'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0138", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,64,Lisbon,hr,131000\n2,ivan,52,Paris,sales,108000\n3,dave,33,Berlin,finance,77000\n4,grace,61,Tokyo,finance,93000\n5,linda,62,Rome,ops,92000\n6,nora,57,Cairo,marketing,99000\n7,carol,30,Rome,finance,57000\n8,judy,36,Rome,sales,145000\n9,mike,36,Tokyo,finance,128000\n10,pete,60,Berlin,sales,103000\n11,olive,37,Rome,ops,43000\n12,alice,38,Lisbon,marketing,138000\n13,frank,25,Madrid,sales,65000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "591", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0139", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,59,London,ops,120000\n2,karl,27,London,sales,59000\n3,dave,64,Paris,hr,133000\n4,carol,27,Paris,eng,104000\n5,mike,63,London,marketing,81000\n6,pete,34,Tokyo,hr,65000\n7,olive,44,Lisbon,sales,86000\n8,judy,27,Berlin,ops,56000\n9,heidi,52,Tokyo,ops,68000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "772000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0140", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,33,Rome,finance,122000\n2,carol,61,Madrid,hr,94000\n3,nora,26,Cairo,eng,122000\n4,linda,52,Berlin,sales,40000\n5,mike,63,Tokyo,eng,93000\n6,frank,22,Tokyo,eng,85000\n7,dave,43,Rome,hr,134000\n8,judy,29,Berlin,eng,147000\n9,pete,23,Madrid,sales,79000\n10,alice,33,Paris,sales,46000\n\n=== bonus_by_dept ===\ndept,bonus\neng,2000\nfinance,5000\nhr,4000\nsales,5000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "998000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0141", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,59,Berlin,ops,66000\n2,ivan,56,Cairo,ops,72000\n3,judy,36,Tokyo,hr,44000\n4,carol,31,Tokyo,hr,136000\n5,eve,30,Cairo,marketing,99000\n6,karl,49,Paris,ops,55000\n7,pete,46,Rome,sales,63000\n8,olive,29,London,marketing,142000\n9,alice,62,London,finance,129000\n10,bob,52,Tokyo,ops,104000\n11,heidi,45,Tokyo,hr,116000\n12,mike,22,Tokyo,marketing,134000\n13,linda,57,Lisbon,marketing,64000\n14,nora,40,Rome,hr,70000\n15,frank,40,Rome,hr,114000\n16,dave,56,London,hr,46000\n```\n\nTask: Group the rows by 'city' and compute the max of 'age' per group. Return the group name with the largest max.", "reference_answer": "London", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0142", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,46,London,eng,42000\n2,karl,38,Paris,eng,63000\n3,mike,41,Berlin,ops,134000\n4,pete,33,Berlin,finance,45000\n5,carol,42,Madrid,sales,108000\n6,grace,57,Lisbon,sales,95000\n7,judy,64,Cairo,marketing,131000\n8,frank,32,London,marketing,122000\n9,nora,62,Berlin,finance,123000\n10,dave,28,Berlin,sales,114000\n11,linda,60,Cairo,marketing,41000\n12,heidi,53,London,hr,140000\n13,alice,39,Madrid,marketing,96000\n14,bob,31,Madrid,marketing,88000\n15,ivan,26,London,ops,133000\n16,olive,33,Lisbon,ops,125000\n17,eve2,31,Madrid,marketing,75000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "hr", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0143", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,56,Paris,marketing,104000\n2,judy,45,Cairo,sales,147000\n3,grace,35,Tokyo,finance,84000\n4,ivan,63,Tokyo,ops,98000\n5,dave,30,Cairo,hr,149000\n6,olive,32,Rome,hr,105000\n7,alice,34,Madrid,ops,100000\n8,eve,21,Berlin,ops,62000\n9,nora,21,Paris,marketing,103000\n10,pete,45,Lisbon,finance,72000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "149000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0144", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,30,Madrid,finance,50000\n2,judy,47,Tokyo,finance,112000\n3,bob,55,Lisbon,finance,110000\n4,olive,53,Cairo,hr,121000\n5,dave,33,Lisbon,finance,58000\n6,nora,49,Tokyo,finance,82000\n7,mike,62,Rome,eng,120000\n8,frank,27,Lisbon,ops,106000\n9,grace,50,London,hr,109000\n10,eve,47,Madrid,eng,69000\n11,pete,22,Rome,marketing,136000\n12,karl,33,Madrid,eng,103000\n13,ivan,50,Berlin,finance,96000\n14,heidi,62,Rome,ops,101000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "48.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0145", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,alice,60,Rome,hr,143000\n2,karl,43,Berlin,eng,74000\n3,linda,46,Madrid,eng,103000\n4,dave,55,Tokyo,finance,64000\n5,nora,36,Rome,ops,104000\n6,bob,36,Madrid,marketing,68000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 4 (0-indexed, header excluded)? Return only the value.", "reference_answer": "36", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0146", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,heidi,46,Lisbon,ops,95000\n2,ivan,37,Lisbon,sales,147000\n3,karl,21,Rome,ops,118000\n4,alice,41,Cairo,ops,61000\n5,pete,44,Tokyo,finance,141000\n6,dave,49,Madrid,ops,82000\n7,olive,37,Cairo,finance,44000\n8,grace,64,Paris,finance,101000\n9,judy,21,Madrid,sales,82000\n10,linda,30,London,sales,116000\n11,bob,32,Paris,sales,146000\n12,frank,35,Tokyo,sales,126000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1259000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0147", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,35,Lisbon,hr,50000\n2,alice,54,Lisbon,eng,45000\n3,heidi,24,Paris,marketing,75000\n4,grace,23,Rome,sales,80000\n5,dave,37,Madrid,marketing,73000\n6,olive,62,Madrid,marketing,50000\n7,nora,45,Madrid,eng,52000\n8,mike,39,Madrid,hr,132000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 3 (0-indexed, header excluded)? Return only the value.", "reference_answer": "23", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0148", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,21,Cairo,marketing,129000\n2,heidi,58,Berlin,eng,111000\n3,judy,46,Cairo,sales,54000\n4,carol,42,Madrid,marketing,47000\n5,karl,55,London,finance,77000\n6,alice,34,Berlin,sales,141000\n7,dave,46,Lisbon,sales,94000\n8,ivan,38,Rome,sales,130000\n9,nora,48,Berlin,eng,56000\n10,grace,37,Berlin,ops,50000\n11,frank,49,Berlin,marketing,109000\n12,linda,35,Rome,eng,78000\n13,pete,39,Tokyo,sales,122000\n```\n\nTask: How many rows have age > 34 AND city = 'Tokyo'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0149", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,pete,53,London,marketing,42000\n2,bob,37,Paris,ops,111000\n3,linda,57,Paris,hr,78000\n4,dave,43,Lisbon,eng,149000\n5,ivan,31,Paris,ops,49000\n6,carol,25,Paris,sales,73000\n7,karl,39,Madrid,ops,108000\n8,olive,37,Madrid,eng,92000\n9,grace,44,Lisbon,marketing,86000\n10,frank,40,London,hr,145000\n\n=== bonus_by_dept ===\ndept,bonus\neng,5000\nhr,2000\nmarketing,8000\nops,7000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "992000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0150", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,53,Rome,finance,67000\n2,linda,21,Tokyo,hr,126000\n3,heidi,60,Rome,finance,73000\n4,pete,64,Lisbon,hr,103000\n5,carol,60,Lisbon,finance,100000\n6,judy,39,Lisbon,finance,68000\n7,dave,53,Cairo,sales,105000\n8,nora,61,Cairo,ops,120000\n9,frank,25,Berlin,eng,129000\n10,karl,58,Berlin,hr,111000\n11,bob,33,London,marketing,85000\n```\n\nTask: In the CSV below, what is the value of column 'salary' in row 5 (0-indexed, header excluded)? Return only the value.", "reference_answer": "68000", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0151", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,24,Madrid,finance,79000\n2,bob,37,Paris,hr,76000\n3,carol,58,Madrid,hr,70000\n4,linda,57,London,ops,118000\n5,karl,42,London,hr,51000\n6,frank,54,Cairo,ops,57000\n7,nora,37,Madrid,eng,53000\n8,mike,58,London,ops,65000\n9,grace,64,Lisbon,sales,56000\n10,heidi,25,Lisbon,sales,134000\n11,dave,50,Tokyo,sales,45000\n12,olive,48,Tokyo,eng,71000\n13,pete,42,Berlin,hr,63000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "48.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0152", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,61,Berlin,eng,41000\n2,nora,52,Madrid,marketing,147000\n3,pete,41,Tokyo,eng,102000\n4,dave,23,London,sales,110000\n5,karl,32,London,sales,142000\n6,bob,47,London,eng,130000\n7,eve,26,Rome,hr,121000\n8,alice,33,Madrid,sales,122000\n9,heidi,36,Cairo,ops,144000\n10,ivan,42,Cairo,ops,42000\n11,grace,43,Lisbon,eng,88000\n12,judy,45,Rome,marketing,111000\n13,carol,59,Madrid,ops,134000\n14,mike,63,Madrid,ops,81000\n15,frank,43,Paris,marketing,96000\n16,linda,50,Rome,eng,66000\n```\n\nTask: Compute the mean of the 'age' column. Return the numeric value (any precision).", "reference_answer": "43.5", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0153", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,56,Tokyo,marketing,56000\n2,grace,35,Lisbon,marketing,81000\n3,karl,53,Tokyo,ops,119000\n4,judy,30,Rome,ops,113000\n5,bob,55,Lisbon,hr,110000\n6,nora,21,Paris,ops,123000\n7,carol,31,Tokyo,finance,122000\n8,pete,42,Berlin,ops,52000\n9,eve,54,Tokyo,marketing,135000\n10,heidi,34,Lisbon,marketing,125000\n11,linda,55,Paris,finance,96000\n12,ivan,26,Paris,marketing,57000\n13,mike,61,Tokyo,finance,109000\n14,olive,39,Paris,ops,83000\n15,alice,59,London,finance,84000\n16,frank,33,Lisbon,marketing,134000\n17,dave2,32,Cairo,finance,51000\n18,grace2,59,Berlin,ops,97000\n19,karl2,28,Tokyo,ops,145000\n20,judy2,60,Lisbon,hr,40000\n21,bob2,30,London,hr,47000\n22,nora2,46,Cairo,sales,137000\n23,carol2,51,London,marketing,73000\n24,pete2,25,Rome,eng,86000\n```\n\nTask: How many rows have age > 44 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0154", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,dave,25,London,finance,77000\n2,carol,47,Paris,ops,46000\n3,linda,41,Berlin,ops,146000\n4,frank,44,London,eng,122000\n5,ivan,46,Cairo,sales,72000\n6,grace,27,London,marketing,85000\n7,alice,31,Madrid,sales,114000\n8,bob,20,Paris,sales,71000\n9,karl,63,Paris,hr,44000\n10,heidi,43,Tokyo,finance,121000\n11,judy,22,Berlin,finance,125000\n12,mike,33,Berlin,eng,138000\n\n=== bonus_by_dept ===\ndept,bonus\neng,1000\nfinance,9000\nhr,5000\nmarketing,3000\nops,1000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1203000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0155", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,55,London,eng,130000\n2,bob,42,London,marketing,94000\n3,heidi,61,Cairo,eng,41000\n4,alice,41,Berlin,finance,133000\n5,olive,48,Madrid,eng,57000\n6,pete,42,Lisbon,eng,98000\n7,mike,47,Paris,finance,128000\n8,linda,53,London,sales,52000\n9,judy,21,Madrid,sales,138000\n10,ivan,57,Cairo,ops,125000\n11,frank,31,Rome,ops,148000\n12,grace,64,Berlin,sales,88000\n13,nora,49,Rome,marketing,82000\n14,karl,47,Berlin,eng,46000\n15,dave,23,Rome,ops,93000\n16,eve,23,Cairo,sales,103000\n17,carol2,56,Berlin,finance,117000\n18,bob2,64,Madrid,sales,136000\n19,heidi2,57,Lisbon,ops,42000\n20,alice2,60,Berlin,eng,44000\n21,olive2,55,Cairo,eng,128000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'salary' per group. Return the group name with the largest mean.", "reference_answer": "Paris", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0156", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,184\n1,189\n2,194\n3,199\n4,204\n5,209\n6,314\n7,219\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "6", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0157", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,35,Berlin,ops,119000\n2,grace,29,Paris,marketing,89000\n3,eve,59,Madrid,finance,121000\n4,pete,49,Rome,eng,61000\n5,karl,61,Cairo,hr,107000\n6,carol,58,Berlin,ops,66000\n7,nora,25,London,hr,103000\n8,olive,39,Berlin,eng,97000\n9,judy,54,Berlin,ops,143000\n10,alice,59,Cairo,hr,68000\n11,dave,48,London,marketing,131000\n12,frank,33,Berlin,hr,50000\n```\n\nTask: Compute the mean of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "96250.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0158", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,27,London,finance,86000\n2,dave,35,London,finance,64000\n3,ivan,45,Cairo,ops,121000\n4,linda,28,Lisbon,eng,115000\n5,bob,31,Berlin,marketing,128000\n6,alice,24,Tokyo,marketing,57000\n7,grace,30,Cairo,sales,115000\n8,eve,36,Tokyo,eng,78000\n9,olive,42,London,sales,102000\n10,nora,53,Tokyo,hr,78000\n11,carol,33,London,ops,113000\n12,judy,31,Rome,sales,148000\n13,pete,23,Berlin,sales,129000\n```\n\nTask: Group the rows by 'dept' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "hr", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0159", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,34,Berlin,eng,144000\n2,mike,54,Rome,sales,62000\n3,pete,30,London,finance,81000\n4,grace,44,Tokyo,marketing,79000\n5,frank,31,Madrid,hr,106000\n6,alice,59,London,finance,83000\n7,nora,35,Rome,sales,51000\n8,dave,35,Berlin,hr,62000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "322", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0160", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,ivan,29,Lisbon,ops,112000\n2,frank,55,Cairo,finance,74000\n3,linda,50,Rome,sales,149000\n4,olive,59,Cairo,finance,114000\n5,alice,54,Paris,finance,132000\n6,heidi,20,Cairo,sales,102000\n7,judy,39,Paris,eng,146000\n8,karl,38,Cairo,sales,110000\n9,mike,20,Lisbon,marketing,107000\n10,nora,38,London,sales,146000\n11,bob,58,Tokyo,marketing,91000\n\n=== bonus_by_dept ===\ndept,bonus\neng,3000\nfinance,3000\nmarketing,7000\nops,5000\nsales,6000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1338000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0161", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,27,Berlin,finance,61000\n2,dave,53,Cairo,ops,66000\n3,heidi,44,Paris,hr,88000\n4,bob,43,Cairo,sales,80000\n5,olive,51,Tokyo,eng,87000\n6,eve,36,London,finance,118000\n7,mike,32,Madrid,ops,100000\n8,karl,51,Cairo,ops,42000\n9,frank,30,Lisbon,finance,147000\n10,grace,23,Tokyo,marketing,79000\n11,judy,40,Tokyo,finance,149000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "430", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0162", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,olive,33,Rome,hr,134000\n2,pete,23,Cairo,eng,59000\n3,dave,44,Lisbon,marketing,131000\n4,carol,39,Madrid,eng,66000\n5,nora,62,Lisbon,sales,46000\n6,judy,52,Rome,sales,86000\n7,mike,58,Rome,marketing,106000\n8,ivan,34,Cairo,hr,140000\n9,grace,23,Paris,marketing,123000\n10,bob,32,Rome,marketing,95000\n11,heidi,48,Cairo,eng,147000\n12,eve,42,London,finance,139000\n13,frank,26,Rome,hr,59000\n14,linda,32,Madrid,ops,50000\n```\n\nTask: How many rows have age > 30 AND city = 'Cairo'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0163", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,62,Lisbon,marketing,86000\n2,frank,60,Lisbon,eng,75000\n3,bob,29,Cairo,sales,105000\n4,olive,62,Lisbon,finance,138000\n5,karl,36,Cairo,eng,54000\n6,alice,57,Lisbon,hr,83000\n7,ivan,42,Rome,finance,49000\n8,judy,49,Cairo,finance,112000\n9,carol,64,Tokyo,marketing,124000\n10,nora,26,Rome,finance,74000\n11,dave,41,London,ops,103000\n12,pete,41,Berlin,hr,128000\n13,linda,62,Lisbon,eng,90000\n```\n\nTask: Compute the median of the 'age' column. Return the numeric value (any precision).", "reference_answer": "49.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0164", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,33,Cairo,eng,103000\n2,mike,41,Lisbon,marketing,49000\n3,eve,42,Berlin,sales,135000\n4,alice,46,Rome,sales,109000\n5,linda,21,Cairo,sales,74000\n6,ivan,23,London,ops,136000\n7,pete,43,London,sales,113000\n8,carol,46,Berlin,hr,123000\n9,frank,50,Paris,ops,113000\n\n=== bonus_by_dept ===\ndept,bonus\neng,2000\nhr,5000\nmarketing,7000\nops,9000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "991000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0165", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,51,Lisbon,ops,78000\n2,pete,35,Tokyo,eng,52000\n3,dave,64,Paris,finance,135000\n4,olive,42,Rome,ops,67000\n5,bob,33,Lisbon,hr,45000\n6,eve,59,Cairo,eng,124000\n7,linda,35,Lisbon,hr,79000\n8,heidi,31,Lisbon,eng,127000\n9,grace,47,Lisbon,finance,138000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "845000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0166", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,164\n1,166\n2,208\n3,170\n4,172\n5,174\n6,176\n7,178\n8,180\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "2", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0167", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,karl,57,Tokyo,hr,41000\n2,frank,43,Tokyo,hr,68000\n3,alice,51,Lisbon,ops,60000\n4,carol,55,Tokyo,finance,123000\n5,olive,26,Cairo,finance,139000\n6,nora,57,Tokyo,marketing,118000\n7,pete,36,Berlin,hr,57000\n8,bob,46,London,sales,45000\n9,grace,43,Lisbon,ops,87000\n10,eve,51,Madrid,marketing,60000\n11,heidi,29,Madrid,sales,81000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,9000\nhr,4000\nmarketing,8000\nops,2000\nsales,9000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "947000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0168", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,42,Paris,finance,121000\n2,karl,42,Cairo,ops,143000\n3,ivan,21,Berlin,ops,66000\n4,grace,48,Madrid,hr,103000\n5,heidi,33,Rome,hr,133000\n6,frank,44,Rome,hr,93000\n7,carol,20,Paris,marketing,69000\n8,olive,38,Paris,hr,80000\n9,pete,25,Tokyo,sales,79000\n10,bob,33,Cairo,hr,75000\n11,eve,23,Paris,eng,105000\n```\n\nTask: Compute the max of the 'salary' column. Return the integer.", "reference_answer": "143000", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "int"}} |
| {"task_id": "data-42-0169", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,judy,52,Madrid,eng,90000\n2,linda,46,Madrid,ops,94000\n3,karl,58,Paris,finance,133000\n4,olive,57,Tokyo,sales,138000\n5,ivan,33,Berlin,finance,146000\n6,grace,21,Madrid,sales,53000\n7,pete,52,Paris,ops,106000\n8,nora,43,Cairo,sales,42000\n9,frank,25,Paris,eng,127000\n10,alice,63,London,ops,66000\n11,bob,52,Paris,hr,122000\n```\n\nTask: In the CSV below, what is the value of column 'dept' in row 2 (0-indexed, header excluded)? Return only the value.", "reference_answer": "finance", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0170", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,64,Madrid,finance,61000\n2,olive,46,Berlin,finance,106000\n3,pete,57,Lisbon,ops,96000\n4,carol,34,London,eng,43000\n5,bob,49,London,sales,72000\n6,mike,40,London,ops,102000\n7,linda,34,Tokyo,marketing,53000\n8,dave,56,Cairo,ops,106000\n9,nora,41,Berlin,marketing,82000\n10,karl,52,Rome,ops,139000\n11,eve,59,Rome,finance,69000\n12,judy,58,Rome,eng,136000\n13,alice,57,London,eng,50000\n14,heidi,57,Paris,finance,86000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "704", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0171", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,eve,53,Tokyo,sales,122000\n2,pete,32,Tokyo,finance,102000\n3,karl,49,Paris,sales,46000\n4,heidi,41,Rome,ops,67000\n5,dave,31,Tokyo,eng,40000\n6,grace,59,Lisbon,finance,131000\n7,frank,38,London,sales,138000\n8,bob,36,Lisbon,sales,59000\n\n=== bonus_by_dept ===\ndept,bonus\neng,4000\nfinance,3000\nops,7000\nsales,8000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "754000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0172", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,21,Madrid,finance,44000\n2,carol,34,London,ops,144000\n3,bob,63,Madrid,marketing,141000\n4,alice,43,Cairo,eng,110000\n5,olive,21,Cairo,hr,44000\n6,frank,36,Lisbon,marketing,136000\n7,heidi,46,Berlin,ops,105000\n8,linda,42,Madrid,hr,142000\n9,karl,27,Rome,sales,46000\n10,ivan,29,Paris,sales,125000\n11,nora,51,Tokyo,finance,55000\n12,grace,55,Lisbon,eng,61000\n13,judy,38,Rome,finance,63000\n```\n\nTask: Group the rows by 'dept' and compute the max of 'salary' per group. Return the group name with the largest max.", "reference_answer": "ops", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0173", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,ivan,27,Cairo,eng,49000\n2,frank,43,London,ops,44000\n3,carol,21,Berlin,eng,97000\n4,linda,23,Tokyo,hr,119000\n5,dave,37,Paris,marketing,77000\n6,nora,63,Lisbon,hr,147000\n7,eve,51,Cairo,eng,60000\n8,alice,63,Cairo,eng,142000\n\n=== bonus_by_dept ===\ndept,bonus\neng,2000\nhr,4000\nmarketing,5000\nops,7000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "763000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0174", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,dave,48,London,marketing,46000\n2,frank,60,Lisbon,eng,115000\n3,ivan,56,Madrid,sales,93000\n4,judy,38,Rome,eng,142000\n5,eve,62,Tokyo,sales,56000\n6,nora,59,Berlin,marketing,140000\n7,heidi,30,Madrid,finance,146000\n8,linda,64,Rome,eng,144000\n9,alice,55,Rome,eng,145000\n10,olive,37,Madrid,eng,107000\n11,pete,61,Berlin,eng,111000\n12,mike,54,Berlin,marketing,115000\n13,karl,64,Tokyo,hr,112000\n14,grace,52,London,eng,122000\n15,bob,21,Paris,marketing,149000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "761", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0175", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,alice,28,London,hr,87000\n2,linda,53,Rome,sales,106000\n3,dave,33,Madrid,ops,85000\n4,ivan,26,Rome,sales,120000\n5,heidi,55,Tokyo,ops,136000\n6,bob,59,London,hr,50000\n7,grace,55,Berlin,finance,125000\n8,eve,27,Madrid,ops,136000\n9,mike,24,Tokyo,marketing,52000\n10,pete,61,Berlin,sales,113000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1010000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0176", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,eve,45,Lisbon,hr,116000\n2,ivan,50,Berlin,finance,55000\n3,grace,44,Paris,eng,100000\n4,frank,53,Madrid,eng,140000\n5,heidi,55,Rome,eng,120000\n6,nora,52,Berlin,eng,121000\n7,karl,60,Tokyo,sales,106000\n8,mike,55,Berlin,sales,125000\n9,pete,51,Tokyo,finance,149000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 5 (0-indexed, header excluded)? Return only the value.", "reference_answer": "Berlin", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0177", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,51,Paris,sales,89000\n2,bob,45,Lisbon,eng,50000\n3,pete,57,Berlin,finance,95000\n4,mike,64,Rome,eng,143000\n5,frank,54,Madrid,eng,88000\n6,karl,40,Paris,marketing,148000\n7,judy,41,Berlin,hr,113000\n8,olive,61,Rome,eng,118000\n9,eve,28,Cairo,marketing,91000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "441", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0178", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,31,London,ops,58000\n2,dave,40,Madrid,ops,74000\n3,carol,58,Tokyo,finance,138000\n4,pete,45,Paris,hr,92000\n5,ivan,64,Berlin,finance,46000\n6,mike,26,Tokyo,eng,108000\n7,nora,49,Berlin,ops,54000\n8,karl,28,Rome,finance,51000\n9,bob,43,Paris,finance,120000\n10,linda,62,Cairo,sales,49000\n\n=== bonus_by_dept ===\ndept,bonus\neng,5000\nfinance,4000\nhr,4000\nops,5000\nsales,7000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "837000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0179", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,grace,60,Paris,finance,83000\n2,olive,54,London,eng,111000\n3,pete,48,Paris,hr,47000\n4,frank,29,London,hr,141000\n5,heidi,46,Berlin,hr,118000\n6,dave,24,London,sales,136000\n7,nora,23,Lisbon,finance,48000\n8,ivan,30,Lisbon,eng,85000\n9,eve,27,London,finance,91000\n10,carol,38,Berlin,marketing,66000\n11,judy,44,Lisbon,marketing,52000\n12,karl,53,London,hr,72000\n13,mike,54,Tokyo,marketing,142000\n\n=== bonus_by_dept ===\ndept,bonus\neng,2000\nfinance,9000\nhr,1000\nmarketing,9000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1255000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0180", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,bob,60,London,eng,110000\n2,alice,50,Paris,ops,106000\n3,nora,63,Cairo,sales,94000\n4,frank,32,Tokyo,hr,68000\n5,eve,40,Cairo,hr,144000\n6,heidi,21,Madrid,eng,141000\n7,judy,22,Tokyo,eng,131000\n8,linda,60,London,sales,139000\n9,pete,31,Rome,ops,115000\n10,dave,37,Cairo,finance,87000\n```\n\nTask: In the CSV below, what is the value of column 'age' in row 2 (0-indexed, header excluded)? Return only the value.", "reference_answer": "63", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0181", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,51,Lisbon,hr,73000\n2,karl,61,Berlin,eng,128000\n3,linda,33,Berlin,sales,42000\n4,heidi,51,Tokyo,marketing,120000\n5,frank,53,London,ops,44000\n6,alice,54,Paris,hr,63000\n7,mike,24,Cairo,marketing,61000\n8,grace,48,Tokyo,finance,131000\n9,olive,37,Madrid,eng,129000\n10,ivan,22,London,sales,131000\n11,nora,62,Rome,sales,90000\n12,judy,41,Paris,sales,65000\n13,bob,39,Madrid,marketing,110000\n14,dave,41,Rome,ops,53000\n15,carol,62,Tokyo,sales,87000\n```\n\nTask: Compute the median of the 'salary' column. Return the numeric value (any precision).", "reference_answer": "87000.0", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0182", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,pete,48,Rome,sales,133000\n2,nora,62,Berlin,marketing,78000\n3,heidi,30,Cairo,sales,110000\n4,ivan,55,Rome,hr,58000\n5,alice,42,Cairo,finance,76000\n6,olive,54,Madrid,eng,136000\n7,bob,50,Madrid,sales,89000\n8,judy,34,Cairo,eng,129000\n9,eve,44,Lisbon,marketing,119000\n10,dave,37,Cairo,hr,115000\n11,linda,32,London,marketing,84000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "488", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0183", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,51,Cairo,marketing,72000\n2,bob,61,Madrid,finance,129000\n3,olive,20,Berlin,hr,68000\n4,alice,63,Rome,eng,124000\n5,frank,54,Paris,sales,47000\n6,dave,32,Berlin,ops,56000\n7,judy,35,Madrid,finance,78000\n8,heidi,44,London,marketing,112000\n9,eve,63,Rome,sales,62000\n10,karl,51,Madrid,ops,51000\n11,mike,22,Tokyo,ops,101000\n12,linda,32,Cairo,ops,95000\n13,carol,55,Cairo,sales,65000\n14,ivan,62,London,hr,99000\n15,pete,28,Cairo,eng,135000\n16,nora,35,Madrid,sales,105000\n17,grace2,35,Berlin,ops,101000\n18,bob2,59,Lisbon,sales,117000\n19,olive2,64,Madrid,eng,47000\n20,alice2,39,Berlin,finance,84000\n21,frank2,29,Cairo,hr,97000\n```\n\nTask: How many rows have age > 29 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "1", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0184", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,karl,62,Cairo,ops,87000\n2,pete,64,Madrid,ops,53000\n3,judy,53,Berlin,eng,50000\n4,heidi,49,Rome,ops,108000\n5,olive,64,Tokyo,finance,70000\n6,bob,53,Lisbon,eng,134000\n7,grace,60,Paris,marketing,122000\n```\n\nTask: In the CSV below, what is the value of column 'salary' in row 3 (0-indexed, header excluded)? Return only the value.", "reference_answer": "108000", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "int"}} |
| {"task_id": "data-42-0185", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,53,Berlin,ops,112000\n2,carol,21,London,ops,69000\n3,alice,58,Tokyo,sales,84000\n4,frank,51,Madrid,hr,147000\n5,ivan,42,London,marketing,51000\n6,linda,58,Paris,hr,43000\n7,eve,36,Rome,eng,48000\n8,judy,50,Cairo,eng,86000\n9,pete,30,Rome,marketing,48000\n10,karl,46,Rome,sales,90000\n11,bob,44,Cairo,finance,80000\n12,grace,64,Cairo,marketing,76000\n13,nora,20,Berlin,hr,118000\n14,olive,28,Madrid,finance,65000\n15,dave,58,Rome,finance,110000\n16,heidi,46,Cairo,finance,73000\n```\n\nTask: How many rows have age > 48 AND city = 'Cairo'? Return the integer count.", "reference_answer": "2", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0186", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,grace,52,Madrid,sales,141000\n2,frank,33,London,eng,96000\n3,bob,30,Cairo,sales,114000\n4,dave,37,Paris,marketing,59000\n5,karl,31,London,marketing,142000\n6,linda,61,Lisbon,marketing,47000\n7,judy,42,Madrid,marketing,128000\n8,carol,46,Berlin,eng,86000\n9,eve,25,Paris,finance,43000\n10,mike,46,Madrid,ops,43000\n11,pete,24,Paris,hr,144000\n12,ivan,22,Rome,marketing,147000\n13,olive,40,Madrid,ops,86000\n14,nora,29,Madrid,eng,70000\n```\n\nTask: Group the rows by 'city' and compute the mean of 'age' per group. Return the group name with the largest mean.", "reference_answer": "Lisbon", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0187", "family": "data", "difficulty": 0.7, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,22,Rome,finance,100000\n2,carol,20,Lisbon,ops,131000\n3,nora,53,Paris,marketing,71000\n4,bob,56,Lisbon,sales,136000\n5,dave,22,Berlin,eng,50000\n6,linda,27,Paris,hr,118000\n7,mike,33,Cairo,sales,66000\n8,olive,31,Madrid,finance,115000\n9,eve,58,Cairo,sales,96000\n10,frank,31,Berlin,eng,86000\n11,heidi,46,Lisbon,finance,134000\n12,karl,50,Paris,sales,55000\n13,alice,39,Lisbon,hr,42000\n14,judy,57,Lisbon,ops,66000\n15,grace,55,Madrid,sales,125000\n16,pete,59,Cairo,finance,48000\n17,ivan2,31,Paris,finance,73000\n18,carol2,59,Berlin,hr,40000\n19,nora2,24,Berlin,sales,87000\n20,bob2,22,Lisbon,marketing,134000\n```\n\nTask: Group the rows by 'city' and compute the max of 'salary' per group. Return the group name with the largest max.", "reference_answer": "Lisbon", "estimated_seconds": 67.0, "metadata": {"generator": "_gen_t4", "tier": 0.7, "kind": "string"}} |
| {"task_id": "data-42-0188", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,frank,52,London,sales,53000\n2,alice,35,Paris,hr,41000\n3,olive,24,Berlin,eng,84000\n4,bob,31,Lisbon,ops,105000\n5,grace,20,Cairo,marketing,142000\n6,heidi,50,Berlin,finance,101000\n7,pete,20,London,sales,148000\n8,ivan,22,Madrid,eng,64000\n9,linda,51,Cairo,hr,140000\n10,mike,59,Tokyo,hr,133000\n\n=== bonus_by_dept ===\ndept,bonus\neng,5000\nfinance,7000\nhr,6000\nmarketing,2000\nops,5000\nsales,7000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "1067000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0189", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,bob,60,Rome,ops,145000\n2,linda,53,Lisbon,finance,137000\n3,dave,43,Lisbon,finance,71000\n4,carol,33,Paris,eng,58000\n5,nora,35,Rome,hr,144000\n6,heidi,34,Rome,marketing,59000\n7,olive,21,London,eng,59000\n8,judy,45,Madrid,finance,103000\n9,alice,42,Rome,marketing,105000\n10,frank,42,Cairo,finance,73000\n11,grace,21,Paris,hr,133000\n12,eve,32,London,hr,128000\n13,ivan,58,Paris,finance,115000\n14,pete,52,Tokyo,eng,92000\n```\n\nTask: Compute the mean of the 'age' column. Return the numeric value (any precision).", "reference_answer": "40.785714285714285", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0190", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,ivan,20,Cairo,eng,122000\n2,linda,51,London,marketing,61000\n3,frank,50,Paris,hr,48000\n4,carol,52,Lisbon,marketing,76000\n5,nora,63,Cairo,sales,147000\n6,olive,43,London,ops,124000\n7,eve,32,London,eng,56000\n8,judy,55,Berlin,marketing,90000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "366", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0191", "family": "data", "difficulty": 0.95, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nt,value\n0,175\n1,178\n2,181\n3,184\n4,187\n5,190\n6,193\n7,196\n8,199\n9,142\n10,205\n```\n\nTask: The CSV below is a time series that is supposed to follow a roughly linear trend. Exactly one data point is an outlier. Return the index (value of column 't') of the outlier.", "reference_answer": "9", "estimated_seconds": 82.0, "metadata": {"generator": "_gen_t6", "tier": 0.95, "kind": "int"}} |
| {"task_id": "data-42-0192", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,mike,58,Berlin,marketing,93000\n2,pete,45,Berlin,marketing,84000\n3,linda,40,Lisbon,marketing,58000\n4,judy,53,Berlin,hr,149000\n5,alice,42,Paris,sales,48000\n6,heidi,36,Madrid,eng,146000\n7,olive,29,London,hr,143000\n8,ivan,32,Lisbon,finance,41000\n9,carol,33,Tokyo,eng,111000\n10,bob,58,London,marketing,67000\n11,eve,54,Berlin,eng,87000\n12,frank,45,London,eng,53000\n13,nora,31,Cairo,marketing,56000\n14,grace,23,Paris,eng,142000\n15,karl,36,Paris,finance,107000\n```\n\nTask: Sum the values in the 'age' column. Return the integer total.", "reference_answer": "615", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0193", "family": "data", "difficulty": 0.25, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,linda,62,Paris,marketing,63000\n2,grace,60,Berlin,eng,120000\n3,alice,42,Lisbon,ops,47000\n4,dave,36,Cairo,finance,133000\n5,mike,42,Lisbon,eng,148000\n6,bob,20,Cairo,finance,83000\n7,pete,41,Berlin,hr,78000\n8,carol,42,Cairo,marketing,119000\n9,heidi,41,Rome,hr,57000\n10,karl,58,Madrid,sales,121000\n11,olive,54,Madrid,hr,45000\n```\n\nTask: Sum the values in the 'salary' column. Return the integer total.", "reference_answer": "1014000", "estimated_seconds": 40.0, "metadata": {"generator": "_gen_t1", "tier": 0.25, "kind": "int"}} |
| {"task_id": "data-42-0194", "family": "data", "difficulty": 0.1, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,frank,45,Cairo,ops,130000\n2,grace,58,Berlin,marketing,89000\n3,judy,25,London,sales,46000\n4,mike,47,Tokyo,sales,79000\n5,dave,48,Paris,ops,146000\n6,alice,21,Cairo,eng,123000\n7,ivan,46,Lisbon,marketing,106000\n8,karl,42,Cairo,sales,127000\n9,pete,20,Tokyo,marketing,80000\n```\n\nTask: In the CSV below, what is the value of column 'city' in row 7 (0-indexed, header excluded)? Return only the value.", "reference_answer": "Cairo", "estimated_seconds": 31.0, "metadata": {"generator": "_gen_t0", "tier": 0.1, "kind": "string"}} |
| {"task_id": "data-42-0195", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,nora,64,Lisbon,finance,116000\n2,carol,33,Tokyo,eng,63000\n3,mike,28,Lisbon,hr,49000\n4,ivan,46,Cairo,eng,120000\n5,olive,34,Rome,finance,90000\n6,eve,61,London,sales,77000\n7,heidi,58,Tokyo,finance,96000\n8,judy,44,Tokyo,marketing,74000\n9,karl,26,Rome,ops,98000\n10,frank,55,Madrid,sales,48000\n11,bob,56,Tokyo,ops,99000\n\n=== bonus_by_dept ===\ndept,bonus\neng,8000\nfinance,6000\nhr,8000\nmarketing,8000\nops,8000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "998000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0196", "family": "data", "difficulty": 0.4, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,32,London,marketing,72000\n2,frank,60,Lisbon,marketing,42000\n3,bob,47,Madrid,hr,81000\n4,linda,46,Paris,marketing,50000\n5,olive,44,Tokyo,sales,41000\n6,karl,32,London,marketing,128000\n7,nora,24,Tokyo,hr,45000\n8,eve,38,Madrid,hr,60000\n9,grace,63,Tokyo,sales,145000\n10,ivan,29,Cairo,ops,78000\n```\n\nTask: Compute the mean of the 'age' column. Return the numeric value (any precision).", "reference_answer": "41.5", "estimated_seconds": 49.0, "metadata": {"generator": "_gen_t2", "tier": 0.4, "kind": "numeric"}} |
| {"task_id": "data-42-0197", "family": "data", "difficulty": 0.55, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\nid,name,age,city,dept,salary\n1,carol,38,London,eng,74000\n2,bob,63,Tokyo,eng,142000\n3,karl,24,Rome,finance,73000\n4,olive,38,Cairo,marketing,84000\n5,linda,58,Berlin,sales,69000\n6,eve,24,Madrid,marketing,145000\n7,dave,44,Rome,hr,95000\n8,alice,42,Berlin,marketing,104000\n9,heidi,42,Berlin,ops,51000\n10,ivan,39,London,sales,81000\n11,grace,42,Cairo,sales,141000\n12,judy,37,Cairo,ops,96000\n13,frank,48,Cairo,eng,93000\n14,nora,46,London,eng,138000\n15,pete,33,London,sales,132000\n16,mike,38,Cairo,ops,81000\n17,carol2,30,Berlin,ops,54000\n18,bob2,23,London,finance,128000\n19,karl2,60,Berlin,marketing,82000\n20,olive2,21,Cairo,hr,139000\n21,linda2,40,Madrid,ops,124000\n22,eve2,56,Tokyo,hr,91000\n23,dave2,20,Paris,ops,68000\n24,alice2,26,Lisbon,eng,73000\n```\n\nTask: How many rows have age > 35 AND city = 'Lisbon'? Return the integer count.", "reference_answer": "0", "estimated_seconds": 58.0, "metadata": {"generator": "_gen_t3", "tier": 0.55, "kind": "int"}} |
| {"task_id": "data-42-0198", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,karl,55,London,eng,103000\n2,nora,30,Berlin,marketing,132000\n3,grace,54,Paris,marketing,59000\n4,eve,49,Tokyo,eng,41000\n5,bob,57,Paris,hr,144000\n6,alice,20,Madrid,finance,80000\n7,dave,52,Paris,eng,45000\n8,judy,35,Lisbon,sales,141000\n9,olive,50,Tokyo,eng,123000\n\n=== bonus_by_dept ===\ndept,bonus\neng,8000\nfinance,3000\nhr,9000\nmarketing,3000\nsales,1000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "919000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
| {"task_id": "data-42-0199", "family": "data", "difficulty": 0.85, "prompt": "You are a data-analysis agent. Below is one or more CSV table(s). Read the data carefully and answer the task. Return only the answer value \u2014 no prose, no commentary.\n\n```csv\n=== employees ===\nid,name,age,city,dept,salary\n1,ivan,45,Tokyo,hr,84000\n2,olive,24,Tokyo,finance,82000\n3,mike,48,Cairo,sales,126000\n4,nora,46,Madrid,marketing,69000\n5,grace,43,Lisbon,sales,139000\n6,alice,33,Lisbon,hr,51000\n7,karl,21,Lisbon,finance,69000\n8,carol,30,Tokyo,finance,131000\n9,eve,51,Tokyo,finance,132000\n\n=== bonus_by_dept ===\ndept,bonus\nfinance,8000\nhr,9000\nmarketing,9000\nsales,9000\n```\n\nTask: You are given two CSVs. Join them on the 'dept' column and compute the sum of (salary + bonus) across all employees. Return the integer.", "reference_answer": "960000", "estimated_seconds": 76.0, "metadata": {"generator": "_gen_t5", "tier": 0.85, "kind": "int"}} |
