new daily data and new mech field for tools

active_traders.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f1a330f2a848851b0ccdcbfd02df84e1b0cd01cb0a9c19fa731f78f98d863ad
-size 81855
+oid sha256:d85374e556117a03039308754ec4187d8afc93a68a33b2afe856c745970bb105
+size 83808

all_trades_profitability.parquet.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6427dbd1315df12c34b6b3ec530911bf49775dd85ec8ffce8d6b481f346caaf
-size 8969394
+oid sha256:f95e9c4362ae7d94282fe31184a931f07ac64d9ebcf6b120a55b109a06b7ec8f
+size 8757914

closed_market_metrics.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:749b0a4be0d2b7238d3665ec2c4452eff326dcc6466cee32a44c6a853c4ae5b2
-size 129119
+oid sha256:f16a737272d9e50721b01f829144fe8c5868209b40e29cb1a2d9ada2fed43114
+size 129617

closed_markets_div.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b4a44ddfa26ef822e6014376e028af8793725aa03a2e0d70fcb50f3c6f9db854
-size 87049
+oid sha256:cb9f5f47c5d3cdeb20a7a2e5925ceaae90c1b7402960889f2e2be01d1742887b
+size 88059

daily_info.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69a36fa8d368eb15ad1e917b52f419a63b35c3228ef24f3fc12b65817903bc05
-size 2124692
+oid sha256:f4a2cf63954ccfebf9bf15019b6c669f7a7877ba39ab5a153c4a0e891d504e01
+size 1846940

daily_mech_requests.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dbb80628ea75af2b93b65326a8d02145d3262c3288fca33661a482a68083267d
-size 10354
+oid sha256:b26a59996fc1f4551f7eb699277bc20fb34a9b3e0288cfbd80d185157a654c8e
+size 10433

error_by_markets.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb2ca33318018c7dd6e723dd7bc093e042710fad8091c47f4f1ffbe30bc2ab10
-size 11583
+oid sha256:8be709dee296b34c00ee9b910d9c387346a833f23da50ae3187a46207554ac3a
+size 11880

invalid_trades.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3791653969f9bc69c14458dc23192ca78b14b367f971338a91ebba3bbd443b3
-size 127404
+oid sha256:147c82a40ceea1593936561f3c277b7b668e3e1e0c662d88613f46f30f7b8402
+size 125231

retention_activity.parquet.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aaffc3062ee8edce7d0d50e73141a783fc85ea15c92c3693bc4054b20511e320
-size 8528662
+oid sha256:8bd869ba65faeeb61b84c4236f3acd6ba172898f6fe13438f5944e87d5405dd1
+size 8730982

scripts/get_mech_info.py

@@ -23,6 +23,7 @@ from mech_request_utils import (
     get_ipfs_data,
     merge_json_files,
 )
+import logging
 
 SUBGRAPH_HEADERS = {
     "Accept": "application/json, multipart/mixed",
@@ -149,7 +150,7 @@ def update_all_trades_parquet(new_trades_df: pd.DataFrame) -> pd.DataFrame:
     return merge_df
 
 
-def update_tools_parquet(new_tools_filename: pd.DataFrame):
+def update_tools_parquet(new_tools_filename: str):
     try:
         old_tools_df = pd.read_parquet(TMP_DIR / "tools.parquet")
     except Exception as e:
@@ -162,6 +163,13 @@ def update_tools_parquet(new_tools_filename: pd.DataFrame):
         print(f"Error reading new trades parquet file {e}")
         return None
 
+    # Add the missing column to old_tools_df with NaN values
+    if (
+        "mech_address" in new_tools_df.columns
+        and "mech_address" not in old_tools_df.columns
+    ):
+        old_tools_df["mech_address"] = None
+
     # merge two dataframes
     merge_df = pd.concat([old_tools_df, new_tools_df], ignore_index=True)
 
@@ -265,7 +273,7 @@ def get_mech_info_last_60_days() -> dict[str, Any]:
 
 
 @measure_execution_time
-def get_mech_events_since_last_run(logger):
+def get_mech_events_since_last_run(logger, mech_sandbox: bool = False):
     """Function to download only the new events since the last execution."""
 
     # Read the latest date from stored data
@@ -297,39 +305,49 @@ def get_mech_events_since_last_run(logger):
     last_block_number = get_last_block_number()
 
     # mech requests
+    req_output_file = "new_mech_requests.json"
+
     requests_dict, duplicatedReqId, nr_errors = collect_all_mech_requests(
         from_block=last_run_block_number,
         to_block=last_block_number,
-        filename="new_mech_requests.json",
+        filename=req_output_file,
+        mech_sandbox=mech_sandbox,
     )
+
     print(f"NUMBER OF MECH REQUEST ERRORS={nr_errors}")
     # mech delivers
+    deli_output_file = "new_mech_delivers.json"
+
     delivers_dict, duplicatedIds, nr_errors = collect_all_mech_delivers(
         from_block=last_run_block_number,
         to_block=last_block_number,
-        filename="new_mech_delivers.json",
+        filename=deli_output_file,
+        mech_sandbox=mech_sandbox,
     )
+
     print(f"NUMBER OF MECH DELIVER ERRORS={nr_errors}")
     if delivers_dict is None:
         return None
     # clean delivers
-    clean_mech_delivers("new_mech_requests.json", "new_mech_delivers.json")
+    clean_mech_delivers(req_output_file, deli_output_file)
 
     # solve duplicated requestIds
-    block_map = fix_duplicate_requestIds(
-        "new_mech_requests.json", "new_mech_delivers.json"
-    )
+    block_map = fix_duplicate_requestIds(req_output_file, deli_output_file)
+
+    merged_output_file = "new_merged_requests.json"
+    tools_output_file = "new_tools_info.json"
     # merge the two files into one source
     not_found = merge_requests_delivers(
-        "new_mech_requests.json", "new_mech_delivers.json", "new_merged_requests.json"
+        req_output_file, deli_output_file, merged_output_file
     )
 
     # Add ipfs contents
-    get_ipfs_data("new_merged_requests.json", "new_tools_info.json", logger)
+    get_ipfs_data(merged_output_file, tools_output_file, logger)
     return latest_timestamp
 
 
 if __name__ == "__main__":
-    get_mech_events_since_last_run()
+    logger = logging.getLogger(__name__)
+    get_mech_events_since_last_run(logger, mech_sandbox=True)
     # result = get_mech_info_last_60_days()
     # print(result)

scripts/mech_request_utils.py

@@ -54,7 +54,9 @@ BLOCKS_CHUNK_SIZE = 10000
 TEXT_ALIGNMENT = 30
 MINIMUM_WRITE_FILE_DELAY_SECONDS = 20
 MECH_FROM_BLOCK_RANGE = 50000
-
+MECH_SANDBOX_SUBGRAPH = (
+    "https://api.studio.thegraph.com/query/81754/mech-sandbox/version/latest"
+)
 last_write_time = 0.0
 
 REQUESTS_QUERY_FILTER = """
@@ -70,6 +72,20 @@ query requests_query($sender_not_in: [Bytes!], $id_gt: Bytes, $blockNumber_gte:
   }
 }
 """
+REQUESTS_BY_MECH_QUERY_FILTER = """
+query requests_query($sender_not_in: [Bytes!], $id_gt: Bytes, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
+  requests(where: {sender_not_in: $sender_not_in, id_gt: $id_gt, blockNumber_gte: $blockNumber_gte, blockNumber_lte: $blockNumber_lte},  orderBy: id, first: 1000) {
+    blockNumber
+    blockTimestamp
+    id
+    ipfsHash
+    requestId
+    sender
+    mech
+    transactionHash
+  }
+}
+"""
 
 DELIVERS_QUERY_NO_FILTER = """
 query delivers_query($id_gt: Bytes, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
@@ -83,8 +99,23 @@ query delivers_query($id_gt: Bytes, $blockNumber_gte: BigInt, $blockNumber_lte:
     transactionHash
   }
 }
+"""
 
+DELIVERS_BY_MECH_QUERY_NO_FILTER = """
+query delivers_query($id_gt: Bytes, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
+  delivers(where: {id_gt: $id_gt, blockNumber_gte: $blockNumber_gte, blockNumber_lte: $blockNumber_lte}, orderBy: id, first: 1000) {
+    blockNumber
+    blockTimestamp
+    id
+    ipfsHash
+    requestId
+    sender
+    mech
+    transactionHash
+  }
+}
 """
+
 DELIVERS_QUERY = """
 query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
   delivers(where: {requestId: $requestId, blockNumber_gte: $blockNumber_gte, blockNumber_lte: $blockNumber_lte}, orderBy: blockNumber, first: 1000) {
@@ -98,6 +129,20 @@ query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_
   }
 }
 """
+DELIVERS_BY_MECH_QUERY = """
+query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
+  delivers(where: {requestId: $requestId, blockNumber_gte: $blockNumber_gte, blockNumber_lte: $blockNumber_lte}, orderBy: blockNumber, first: 1000) {
+    blockNumber
+    blockTimestamp
+    id
+    ipfsHash
+    requestId
+    sender
+    mech
+    transactionHash
+  }
+}
+"""
 
 MISSING_DELIVERS_QUERY = """
 query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
@@ -114,12 +159,37 @@ query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_
 """
 
 
-def collect_all_mech_requests(from_block: int, to_block: int, filename: str) -> Tuple:
+MISSING_BY_MECH_DELIVERS_QUERY = """
+query delivers_query($requestId: BigInt, $blockNumber_gte: BigInt, $blockNumber_lte: BigInt) {
+  delivers(where: {requestId: $requestId, blockNumber_gte: $blockNumber_gte, blockNumber_lte: $blockNumber_lte}, orderBy: blockNumber, first: 1000) {
+    blockNumber
+    blockTimestamp
+    id
+    ipfsHash
+    requestId
+    sender
+    mech
+    transactionHash
+  }
+}
+"""
+
+
+def collect_all_mech_requests(
+    from_block: int, to_block: int, filename: str, mech_sandbox: bool = False
+) -> Tuple:
 
     print(f"Fetching all mech requests from {from_block} to {to_block}")
     mech_requests = {}
     duplicated_reqIds = []
-    mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(subgraph_api_key=SUBGRAPH_API_KEY)
+    if mech_sandbox:
+        mech_subgraph_url = MECH_SANDBOX_SUBGRAPH
+        request_query = REQUESTS_BY_MECH_QUERY_FILTER
+    else:
+        mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(
+            subgraph_api_key=SUBGRAPH_API_KEY
+        )
+        request_query = REQUESTS_QUERY_FILTER
     transport = RequestsHTTPTransport(url=mech_subgraph_url)
     client = Client(transport=transport, fetch_schema_from_transport=True)
 
@@ -133,7 +203,7 @@ def collect_all_mech_requests(from_block: int, to_block: int, filename: str) ->
             "blockNumber_lte": str(to_block),  # str
         }
         try:
-            response = fetch_with_retry(client, REQUESTS_QUERY_FILTER, variables)
+            response = fetch_with_retry(client, request_query, variables)
 
             items = response.get("requests", [])
 
@@ -174,13 +244,23 @@ def fetch_with_retry(client, query, variables, max_retries=5):
             time.sleep(wait_time)
 
 
-def collect_all_mech_delivers(from_block: int, to_block: int, filename: str) -> Tuple:
+def collect_all_mech_delivers(
+    from_block: int, to_block: int, filename: str, mech_sandbox: bool = False
+) -> Tuple:
 
     print(f"Fetching all mech delivers from {from_block} to {to_block}")
 
     mech_delivers = {}
     duplicated_requestIds = []
-    mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(subgraph_api_key=SUBGRAPH_API_KEY)
+    if mech_sandbox:
+        mech_subgraph_url = MECH_SANDBOX_SUBGRAPH
+        deliver_query = DELIVERS_BY_MECH_QUERY_NO_FILTER
+    else:
+        mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(
+            subgraph_api_key=SUBGRAPH_API_KEY
+        )
+        deliver_query = DELIVERS_QUERY_NO_FILTER
+
     transport = RequestsHTTPTransport(url=mech_subgraph_url)
     client = Client(transport=transport, fetch_schema_from_transport=True)
     to_block = (
@@ -195,7 +275,7 @@ def collect_all_mech_delivers(from_block: int, to_block: int, filename: str) ->
             "blockNumber_lte": str(to_block),  # str
         }
         try:
-            response = fetch_with_retry(client, DELIVERS_QUERY_NO_FILTER, variables)
+            response = fetch_with_retry(client, deliver_query, variables)
             items = response.get("delivers", [])
 
             if not items:
@@ -225,13 +305,23 @@ def collect_all_mech_delivers(from_block: int, to_block: int, filename: str) ->
     return mech_delivers, duplicated_requestIds, nr_errors
 
 
-def collect_missing_delivers(request_id: int, block_number: int) -> Dict[str, Any]:
+def collect_missing_delivers(
+    request_id: int, block_number: int, mech_sandbox: bool = False
+) -> Dict[str, Any]:
     to_block = (
         block_number + MECH_FROM_BLOCK_RANGE
     )  # there is a delay between deliver and request
     print(f"Fetching all missing delivers from {block_number} to {to_block}")
     mech_delivers = {}
-    mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(subgraph_api_key=SUBGRAPH_API_KEY)
+
+    if mech_sandbox:
+        mech_subgraph_url = MECH_SANDBOX_SUBGRAPH
+        missing_query = MISSING_BY_MECH_DELIVERS_QUERY
+    else:
+        mech_subgraph_url = MECH_SUBGRAPH_URL.substitute(
+            subgraph_api_key=SUBGRAPH_API_KEY
+        )
+        missing_query = MISSING_DELIVERS_QUERY
     transport = RequestsHTTPTransport(url=mech_subgraph_url)
     client = Client(transport=transport, fetch_schema_from_transport=True)
 
@@ -241,7 +331,7 @@ def collect_missing_delivers(request_id: int, block_number: int) -> Dict[str, An
         "blockNumber_lte": str(to_block),  # str
     }
     try:
-        response = fetch_with_retry(client, MISSING_DELIVERS_QUERY, variables)
+        response = fetch_with_retry(client, missing_query, variables)
         items = response.get("delivers", [])
         # If the user sends requests with the same values (tool, prompt, nonce) it
         # will generate the same requestId. Therefore, multiple items can be retrieved

scripts/pull_data.py

@@ -96,7 +96,7 @@ def only_new_weekly_analysis():
     # Mech events ETL
     logging.info("Generating the mech json files")
     # get only new data
-    latest_timestamp = get_mech_events_since_last_run(logger)
+    latest_timestamp = get_mech_events_since_last_run(logger, mech_sandbox=True)
     if latest_timestamp == None:
         print("Error while getting the mech events")
         return
@@ -135,7 +135,7 @@ def only_new_weekly_analysis():
 
     save_historical_data()
     try:
-        clean_old_data_from_parquet_files("2025-01-23")
+        clean_old_data_from_parquet_files("2025-01-26")
     except Exception as e:
         print("Error cleaning the oldest information from parquet files")
         print(f"reason = {e}")
@@ -169,8 +169,8 @@ def restoring_trades_data(from_date: str, to_date: str):
 
 
 if __name__ == "__main__":
-    # only_new_weekly_analysis()
-    save_historical_data()
+    only_new_weekly_analysis()
+    # save_historical_data()
     # try:
     #     clean_old_data_from_parquet_files("2025-01-23")
     # except Exception as e:

scripts/tools.py

@@ -208,7 +208,7 @@ def parse_json_events(json_events: dict, keys_to_traverse: List[int]) -> pd.Data
             output["error"] = error_value
             output["error_message"] = error_message_value
             output["prompt_response"] = json_input["deliver"]["ipfsContents"]["prompt"]
-            output["mech_address"] = json_input["deliver"]["sender"]
+            output["mech_address"] = json_input["deliver"]["mech"]
             p_yes_value, p_no_value, confidence_value, info_utility_value = (
                 get_prediction_values(prediction_params)
             )
@@ -317,4 +317,7 @@ def generate_tools_file(input_filename: str, output_filename: str):
 
 if __name__ == "__main__":
 
-    generate_tools_file()
+    generate_tools_file(
+        input_filename="mechsandbox_tools_info.json",
+        output_filename="mechsandbox_tools.parquet",
+    )

service_map.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77b4d92c14485e09f270c2538742ac508f345abdd225be796a213b212213c1ab
-size 193777
+oid sha256:49a190affe8e9179edd8c80fb249b91c608853f2ef1df83735d3b25d1cac92e6
+size 200308

tools_accuracy.csv

@@ -1,13 +1,13 @@
 tool,tool_accuracy,total_requests,min,max
-claude-prediction-offline,63.23015175787642,58053,2025-01-23 00:02:25,2025-03-22 22:25:30
-claude-prediction-online,64.43582942256683,34307,2025-01-23 00:01:15,2025-03-23 01:42:25
-prediction-offline,53.15215124765413,158257,2025-01-23 00:00:20,2025-03-23 17:36:15
-prediction-offline-sme,53.21729365524986,8905,2025-01-23 00:14:45,2025-03-22 23:56:25
-prediction-online,44.550063371356146,4734,2025-01-23 08:01:20,2025-03-23 16:24:35
-prediction-online-sme,43.840451817860924,2833,2025-01-23 08:07:45,2025-03-22 18:15:50
-prediction-request-rag,46.47519582245431,383,2025-01-23 12:21:05,2025-03-21 19:28:50
-prediction-request-rag-claude,48.732394366197184,355,2025-01-23 12:07:25,2025-03-19 15:45:45
-prediction-request-reasoning,49.19531698100476,59278,2025-01-23 09:22:35,2025-03-22 18:27:30
-prediction-request-reasoning-claude,68.01242236024845,322,2025-01-23 09:59:40,2025-03-20 13:33:50
-prediction-url-cot-claude,54.12147505422994,922,2025-01-23 09:31:20,2025-03-21 23:32:05
-superforcaster,57.14131955656011,9291,2025-01-23 00:16:20,2025-03-23 19:24:40
+claude-prediction-offline,63.046214390810306,56411,2025-01-26 00:04:55,2025-03-25 23:13:55
+claude-prediction-online,64.30493815960567,33069,2025-01-26 00:00:05,2025-03-26 01:19:40
+prediction-offline,53.52954942970462,162372,2025-01-26 00:00:40,2025-03-26 23:44:05
+prediction-offline-sme,49.66247518199868,7555,2025-01-26 00:18:20,2025-03-25 18:52:45
+prediction-online,44.43744752308984,4764,2025-01-26 08:19:45,2025-03-26 10:04:25
+prediction-online-sme,42.7536231884058,2898,2025-01-26 08:21:10,2025-03-26 11:04:25
+prediction-request-rag,45.96774193548387,372,2025-01-26 12:30:50,2025-03-25 18:09:15
+prediction-request-rag-claude,49.56011730205279,341,2025-01-26 09:14:15,2025-03-24 17:36:30
+prediction-request-reasoning,48.855616051623315,59508,2025-01-26 07:59:40,2025-03-26 17:14:00
+prediction-request-reasoning-claude,70.60606060606061,330,2025-01-26 12:09:10,2025-03-25 14:52:45
+prediction-url-cot-claude,53.493699885452465,873,2025-01-26 01:08:05,2025-03-25 22:11:55
+superforcaster,56.70432743551481,9382,2025-01-26 01:17:25,2025-03-25 23:12:40

unknown_traders.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aa637b640db3a761e192450351c2cee555863412f76cad3c4b6abfeff9c4db8d
-size 1006605
+oid sha256:25eca09b15491791210e54fb952673c6d3c60658fd45a568ceddd5939f392904
+size 1062580

weekly_mech_calls.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48f406494a32375fc732c7d100a49f407b3bffa950bbb5e5546c2b7a87fc27a7
-size 57176
+oid sha256:0a66cc54ce2c5522727e63439c1be576c68b3283c5580439a693316107c95ceb
+size 58183

winning_df.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2b70b10b47219e31b47a623db398bf83837a68f6b2d8bd1750e96fbabec8aec8
-size 11059
+oid sha256:662aaefbf51f881b8afd84d8523cb2fd39f8212f57570be5e38ee656666a8e50
+size 11398