ruff_errors.txt

��F401 `app.dataset.load_documents` imported but unused; consider removing, adding to `__all__`, or using a redundant alias
 --> app\__init__.py:1:25
  |
1 | from app.dataset import load_documents
  |                         ^^^^^^^^^^^^^^
2 | from app.cache import SemanticCache
  |
help: Use an explicit re-export: `load_documents as load_documents`

F401 `app.cache.SemanticCache` imported but unused; consider removing, adding to `__all__`, or using a redundant alias
 --> app\__init__.py:2:23
  |
1 | from app.dataset import load_documents
2 | from app.cache import SemanticCache
  |                       ^^^^^^^^^^^^^
  |
help: Use an explicit re-export: `SemanticCache as SemanticCache`

E402 Module level import not at top of file
  --> app\api.py:13:1
   |
11 | # ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
12 |
13 | from app.rate_limit import limiter
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
14 | from app.logger import get_logger
15 | import time
   |

E402 Module level import not at top of file
  --> app\api.py:14:1
   |
13 | from app.rate_limit import limiter
14 | from app.logger import get_logger
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
15 | import time
16 | from app.analytics import log_query, get_analytics_stats
   |

E402 Module level import not at top of file
  --> app\api.py:15:1
   |
13 | from app.rate_limit import limiter
14 | from app.logger import get_logger
15 | import time
   | ^^^^^^^^^^^
16 | from app.analytics import log_query, get_analytics_stats
   |

E402 Module level import not at top of file
  --> app\api.py:16:1
   |
14 | from app.logger import get_logger
15 | import time
16 | from app.analytics import log_query, get_analytics_stats
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
17 |
18 | logger = get_logger("api")
   |

E402 Module level import not at top of file
   --> app\api.py:133:1
    |
131 | # Endpoints
132 | # ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
133 | from app.auth import require_role, Role
    | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
134 |
135 | @router.post("/query", response_model=QueryResponse, dependencies=[Depends(require_role(Role.USER))])
    |

F811 Redefinition of unused `store` from line 274
   --> app\cache.py:305:9
    |
303 |             The newly created CacheEntry
304 |         """
305 |     def store(self, query: str, embedding: np.ndarray, result: str,
    |         ^^^^^ `store` redefined here
306 |               dominant_cluster: int, cluster_probs: np.ndarray | None = None,
307 |               generated_answer: str | None = None,
    |
   ::: app\cache.py:274:9
    |
272 |     # ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
273 |
274 |     def store(
    |         ----- previous definition of `store` here
275 |         self,
276 |         query: str,
    |
help: Remove definition: `store`

E402 Module level import not at top of file
  --> app\main.py:10:1
   |
 8 | load_dotenv()
 9 |
10 | from app.api import router
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^
11 | from app.dataset import load_documents
12 | from app.embeddings import get_model, embed_documents, save_embeddings, load_embeddings
   |

E402 Module level import not at top of file
  --> app\main.py:11:1
   |
10 | from app.api import router
11 | from app.dataset import load_documents
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
12 | from app.embeddings import get_model, embed_documents, save_embeddings, load_embeddings
13 | from app.vector_store import build_index, save_index, load_index
   |

E402 Module level import not at top of file
  --> app\main.py:12:1
   |
10 | from app.api import router
11 | from app.dataset import load_documents
12 | from app.embeddings import get_model, embed_documents, save_embeddings, load_embeddings
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
13 | from app.vector_store import build_index, save_index, load_index
14 | from app.clustering import (
   |

E402 Module level import not at top of file
  --> app\main.py:13:1
   |
11 | from app.dataset import load_documents
12 | from app.embeddings import get_model, embed_documents, save_embeddings, load_embeddings
13 | from app.vector_store import build_index, save_index, load_index
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
14 | from app.clustering import (
15 |     reduce_dimensions, fit_gmm, get_cluster_distributions,
   |

E402 Module level import not at top of file
  --> app\main.py:14:1
   |
12 |   from app.embeddings import get_model, embed_documents, save_embeddings, load_embeddings
13 |   from app.vector_store import build_index, save_index, load_index
14 | / from app.clustering import (
15 | |     reduce_dimensions, fit_gmm, get_cluster_distributions,
16 | |     get_dominant_cluster, save_clustering, load_clustering,
17 | |     analyze_clusters
18 | | )
   | |_^
19 |   from app.cache import SemanticCache
20 |   from app.hybrid_search import BM25Index, HybridSearcher, save_bm25, load_bm25
   |

E402 Module level import not at top of file
  --> app\main.py:19:1
   |
17 |     analyze_clusters
18 | )
19 | from app.cache import SemanticCache
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
20 | from app.hybrid_search import BM25Index, HybridSearcher, save_bm25, load_bm25
21 | from app.logger import setup_logging, get_logger
   |

E402 Module level import not at top of file
  --> app\main.py:20:1
   |
18 | )
19 | from app.cache import SemanticCache
20 | from app.hybrid_search import BM25Index, HybridSearcher, save_bm25, load_bm25
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
21 | from app.logger import setup_logging, get_logger
22 | from app.rate_limit import setup_rate_limiting
   |

E402 Module level import not at top of file
  --> app\main.py:21:1
   |
19 | from app.cache import SemanticCache
20 | from app.hybrid_search import BM25Index, HybridSearcher, save_bm25, load_bm25
21 | from app.logger import setup_logging, get_logger
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
22 | from app.rate_limit import setup_rate_limiting
23 | from app.analytics import init_db
   |

E402 Module level import not at top of file
  --> app\main.py:22:1
   |
20 | from app.hybrid_search import BM25Index, HybridSearcher, save_bm25, load_bm25
21 | from app.logger import setup_logging, get_logger
22 | from app.rate_limit import setup_rate_limiting
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
23 | from app.analytics import init_db
   |

E402 Module level import not at top of file
  --> app\main.py:23:1
   |
21 | from app.logger import setup_logging, get_logger
22 | from app.rate_limit import setup_rate_limiting
23 | from app.analytics import init_db
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
24 |
25 | # Setup structured logging
   |

E402 Module level import not at top of file
   --> app\main.py:112:1
    |
110 | # ���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
111 |
112 | from fastapi.middleware.cors import CORSMiddleware
    | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
113 |
114 | app = FastAPI(
    |

E701 Multiple statements on one line (colon)
   --> experiments\benchmark_expansion.py:113:24
    |
111 |         pool = set()
112 |         for idx in dense_indices:
113 |             if idx >= 0: pool.add(int(idx))
    |                        ^
114 |         for idx in bm25_indices:
115 |             if idx >= 0: pool.add(int(idx))
    |

E701 Multiple statements on one line (colon)
   --> experiments\benchmark_expansion.py:115:24
    |
113 |             if idx >= 0: pool.add(int(idx))
114 |         for idx in bm25_indices:
115 |             if idx >= 0: pool.add(int(idx))
    |                        ^
116 |         for idx in hybrid_indices:
117 |             if idx >= 0: pool.add(int(idx))
    |

E701 Multiple statements on one line (colon)
   --> experiments\benchmark_expansion.py:117:24
    |
115 |             if idx >= 0: pool.add(int(idx))
116 |         for idx in hybrid_indices:
117 |             if idx >= 0: pool.add(int(idx))
    |                        ^
118 |             
119 |         candidates_to_judge = [doc_id for doc_id in pool if doc_id != baseline_id]
    |

E701 Multiple statements on one line (colon)
  --> experiments\chunking_comparison.py:89:50
   |
87 |             q_text = item["query"]
88 |             expected_cat_name = item["category"]
89 |             if expected_cat_name not in cat_to_id: continue
   |                                                  ^
90 |             
91 |             expected_cat_id = cat_to_id[expected_cat_name]
   |

E701 Multiple statements on one line (colon)
  --> experiments\chunking_comparison.py:94:33
   |
92 |             # Ground truth is documents that match the category
93 |             relevant_docs = [i for i, lbl in enumerate(labels) if lbl == expected_cat_id]
94 |             if not relevant_docs: continue
   |                                 ^
95 |             
96 |             q_emb = model.encode([q_text], convert_to_numpy=True, normalize_embeddings=True)[0]
   |

E701 Multiple statements on one line (colon)
   --> experiments\chunking_comparison.py:104:33
    |
102 |             seen = set()
103 |             for chunk_idx in retrieved_chunk_indices:
104 |                 if chunk_idx < 0: continue
    |                                 ^
105 |                 doc_idx = chunk_to_doc[int(chunk_idx)]
106 |                 if doc_idx not in seen:
    |

F841 Local variable `cat_to_id` is assigned to but never used
  --> experiments\error_analysis.py:57:5
   |
55 |     error_report = []
56 |     
57 |     cat_to_id = {name: i for i, name in enumerate(label_names)}
   |     ^^^^^^^^^
58 |     for item in benchmark_queries:
59 |         q_text = item["query"]
   |
help: Remove assignment to unused variable `cat_to_id`

F821 Undefined name `expected_cat_name`
   --> experiments\error_analysis.py:105:51
    |
103 |             reason = "Reranker pushed irrelevant documents above the relevant ones found by hybrid."
104 |         elif not hybrid_hit and not reranked_hit:
105 |             if set(retrieved_cats_hybrid) == set([expected_cat_name]):
    |                                                   ^^^^^^^^^^^^^^^^^
106 |                  failure_type = "DATASET_LABEL_ISSUE"
107 |                  reason = "Retrieved documents look perfectly relevant but ground truth labels don't match."
    |

F821 Undefined name `expected_cat_name`
   --> experiments\error_analysis.py:110:32
    |
108 |             else:
109 |                  # Check if the query is ambiguous between two categories
110 |                  if "cs.LG" in expected_cat_name and any("cs." in c for c in retrieved_cats_hybrid):
    |                                ^^^^^^^^^^^^^^^^^
111 |                      failure_type = "BOUNDARY_CASE"
112 |                      reason = f"Query is ambiguous between {expected_cat_name} and {set(retrieved_cats_hybrid)}."
    |

F821 Undefined name `expected_cat_name`
   --> experiments\error_analysis.py:112:61
    |
110 |                  if "cs.LG" in expected_cat_name and any("cs." in c for c in retrieved_cats_hybrid):
111 |                      failure_type = "BOUNDARY_CASE"
112 |                      reason = f"Query is ambiguous between {expected_cat_name} and {set(retrieved_cats_hybrid)}."
    |                                                             ^^^^^^^^^^^^^^^^^
113 |                  else:
114 |                      failure_type = "RETRIEVAL_FAILURE"
    |

F821 Undefined name `expected_cat_name`
   --> experiments\error_analysis.py:119:34
    |
117 |         error_report.append({
118 |             "query": q_text,
119 |             "expected_category": expected_cat_name,
    |                                  ^^^^^^^^^^^^^^^^^
120 |             "hybrid_retrieved_categories": retrieved_cats_hybrid,
121 |             "reranked_retrieved_categories": retrieved_cats_reranked,
    |

E722 Do not use bare `except`
  --> experiments\evaluate_rag.py:66:5
   |
64 |         if matches:
65 |             return float(matches[0])
66 |     except:
   |     ^^^^^^
67 |         pass
68 |     return 0.0
   |

F841 Local variable `cat_to_id` is assigned to but never used
  --> experiments\evaluate_search.py:62:5
   |
60 |     }
61 |     
62 |     cat_to_id = {name: i for i, name in enumerate(label_names)}
   |     ^^^^^^^^^
63 |     
64 |     for item in benchmark_queries:
   |
help: Remove assignment to unused variable `cat_to_id`

E701 Multiple statements on one line (colon)
  --> experiments\kaggle_bge_benchmark.ipynb:cell 3:15:25
   |
13 | def mrr(relevant, retrieved):
14 |     for rank, d in enumerate(retrieved, 1):
15 |         if d in relevant: return 1.0 / rank
   |                         ^
16 |     return 0.0
   |

E701 Multiple statements on one line (colon)
  --> experiments\kaggle_bge_benchmark.ipynb:cell 3:27:20
   |
26 | def recall_at_k(relevant, retrieved, k):
27 |     if not relevant: return 0.0
   |                    ^
28 |     hits = len(set(retrieved[:k]).intersection(set(relevant)))
29 |     return hits / len(relevant)
   |

E701 Multiple statements on one line (colon)
  --> experiments\kaggle_bge_benchmark.ipynb:cell 3:32:20
   |
31 | def average_precision(relevant, retrieved):
32 |     if not relevant: return 0.0
   |                    ^
33 |     hits = 0
34 |     sum_prec = 0.0
   |

E701 Multiple statements on one line (colon)
  --> experiments\kaggle_bge_benchmark.ipynb:cell 4:32:36
   |
30 |         scores = np.zeros(self.n_docs, dtype=np.float64)
31 |         for term in tokens:
32 |             if term not in self.idf: continue
   |                                    ^
33 |             idf = self.idf[term]
34 |             for doc_idx in range(self.n_docs):
   |

E701 Multiple statements on one line (colon)
  --> experiments\kaggle_bge_benchmark.ipynb:cell 4:36:27
   |
34 |             for doc_idx in range(self.n_docs):
35 |                 tf = self.doc_term_freqs[doc_idx].get(term, 0)
36 |                 if tf == 0: continue
   |                           ^
37 |                 doc_len = self.doc_lengths[doc_idx]
38 |                 num = tf * (self.k1 + 1)
   |

E402 Module level import not at top of cell
  --> experiments\kaggle_bge_benchmark.ipynb:cell 5:11:1
   |
10 | qrels = load_dataset("mteb/scifact", "default", split="test")
11 | from collections import defaultdict
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
12 | benchmark = defaultdict(list)
13 | for item in qrels:
   |

E402 Module level import not at top of file
  --> tests\conftest.py:19:1
   |
18 | # Inject the mock
19 | import sentence_transformers
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
20 | sentence_transformers.SentenceTransformer = MockSentenceTransformer
   |

E402 Module level import not at top of file
  --> tests\conftest.py:23:1
   |
22 | # Now import the app safely
23 | from app.main import app
   | ^^^^^^^^^^^^^^^^^^^^^^^^
24 |
25 | @pytest.fixture
   |

F841 Local variable `score_jumps` is assigned to but never used
  --> tests\unit\test_bm25.py:41:5
   |
39 |     # "the" is in 2 docs, "jumps" is in 1 doc.
40 |     # We don't have direct access to vocab mapping, but we can score single terms
41 |     score_jumps = bm25.score("jumps")[0]
   |     ^^^^^^^^^^^
42 |     score_the = bm25.score("the")[0]
43 |     # Because jumps is rarer, its IDF is higher, and the doc is longer, but usually IDF dominates
   |
help: Remove assignment to unused variable `score_jumps`

F841 Local variable `score_the` is assigned to but never used
  --> tests\unit\test_bm25.py:42:5
   |
40 |     # We don't have direct access to vocab mapping, but we can score single terms
41 |     score_jumps = bm25.score("jumps")[0]
42 |     score_the = bm25.score("the")[0]
   |     ^^^^^^^^^
43 |     # Because jumps is rarer, its IDF is higher, and the doc is longer, but usually IDF dominates
44 |     # Let's just check that IDF values are positive
   |
help: Remove assignment to unused variable `score_the`

Found 42 errors.
No fixes available (4 hidden fixes can be enabled with the `--unsafe-fixes` option).