churnkit 0.75.0a3__py3-none-any.whl → 0.75.1a1__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (45) hide show
  1. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/00_start_here.ipynb +72 -72
  2. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01_data_discovery.ipynb +134 -134
  3. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01a_temporal_deep_dive.ipynb +207 -207
  4. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01b_temporal_quality.ipynb +109 -109
  5. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01c_temporal_patterns.ipynb +283 -283
  6. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01d_event_aggregation.ipynb +145 -145
  7. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/02_column_deep_dive.ipynb +126 -126
  8. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/03_quality_assessment.ipynb +149 -149
  9. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/04_relationship_analysis.ipynb +172 -172
  10. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/05_multi_dataset.ipynb +130 -130
  11. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/06_feature_opportunities.ipynb +163 -163
  12. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/07_modeling_readiness.ipynb +99 -99
  13. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/08_baseline_experiments.ipynb +126 -126
  14. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/09_business_alignment.ipynb +89 -89
  15. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/10_spec_generation.ipynb +132 -132
  16. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/11_scoring_validation.ipynb +197 -197
  17. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/12_view_documentation.ipynb +27 -27
  18. {churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/METADATA +2 -2
  19. {churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/RECORD +45 -45
  20. customer_retention/__init__.py +1 -1
  21. customer_retention/analysis/business/fairness_analyzer.py +2 -2
  22. customer_retention/analysis/diagnostics/segment_analyzer.py +3 -3
  23. customer_retention/analysis/interpretability/cohort_analyzer.py +4 -4
  24. customer_retention/core/compat/__init__.py +20 -0
  25. customer_retention/stages/features/behavioral_features.py +3 -3
  26. customer_retention/stages/features/customer_segmentation.py +10 -10
  27. customer_retention/stages/features/feature_selector.py +2 -2
  28. customer_retention/stages/profiling/relationship_recommender.py +2 -2
  29. customer_retention/stages/profiling/temporal_feature_analyzer.py +3 -3
  30. customer_retention/stages/profiling/temporal_pattern_analyzer.py +10 -10
  31. customer_retention/stages/profiling/temporal_quality_checks.py +6 -6
  32. customer_retention/stages/profiling/time_series_profiler.py +13 -7
  33. customer_retention/stages/profiling/time_window_aggregator.py +22 -15
  34. customer_retention/stages/transformation/categorical_encoder.py +2 -2
  35. customer_retention/stages/transformation/pipeline.py +2 -2
  36. customer_retention/stages/validation/data_quality_gate.py +5 -5
  37. customer_retention/stages/validation/data_validators.py +3 -3
  38. customer_retention/stages/validation/leakage_gate.py +4 -4
  39. customer_retention/stages/validation/timeseries_detector.py +6 -6
  40. customer_retention/transforms/ops.py +2 -2
  41. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/01a_a_temporal_text_deep_dive.ipynb +0 -0
  42. {churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/02a_text_columns_deep_dive.ipynb +0 -0
  43. {churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/WHEEL +0 -0
  44. {churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/entry_points.txt +0 -0
  45. {churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/licenses/LICENSE +0 -0
{churnkit-0.75.0a3.data → churnkit-0.75.1a1.data}/data/share/churnkit/exploration_notebooks/12_view_documentation.ipynb RENAMED
@@ -5,10 +5,10 @@
5
5
  "id": "cell-0",
6
6
  "metadata": {
7
7
  "papermill": {
8
- "duration": 0.001463,
9
- "end_time": "2026-02-02T13:04:38.753851",
8
+ "duration": 0.002863,
9
+ "end_time": "2026-02-03T01:49:08.644729",
10
10
  "exception": false,
11
- "start_time": "2026-02-02T13:04:38.752388",
11
+ "start_time": "2026-02-03T01:49:08.641866",
12
12
  "status": "completed"
13
13
  },
14
14
  "tags": []
@@ -27,16 +27,16 @@
27
27
  "id": "cell-1",
28
28
  "metadata": {
29
29
  "execution": {
30
- "iopub.execute_input": "2026-02-02T13:04:38.756679Z",
31
- "iopub.status.busy": "2026-02-02T13:04:38.756540Z",
32
- "iopub.status.idle": "2026-02-02T13:04:39.108351Z",
33
- "shell.execute_reply": "2026-02-02T13:04:39.107690Z"
30
+ "iopub.execute_input": "2026-02-03T01:49:08.650417Z",
31
+ "iopub.status.busy": "2026-02-03T01:49:08.650253Z",
32
+ "iopub.status.idle": "2026-02-03T01:49:09.163891Z",
33
+ "shell.execute_reply": "2026-02-03T01:49:09.163221Z"
34
34
  },
35
35
  "papermill": {
36
- "duration": 0.354251,
37
- "end_time": "2026-02-02T13:04:39.109163",
36
+ "duration": 0.517356,
37
+ "end_time": "2026-02-03T01:49:09.164527",
38
38
  "exception": false,
39
- "start_time": "2026-02-02T13:04:38.754912",
39
+ "start_time": "2026-02-03T01:49:08.647171",
40
40
  "status": "completed"
41
41
  },
42
42
  "tags": []
@@ -53,16 +53,16 @@
53
53
  "id": "cell-2",
54
54
  "metadata": {
55
55
  "execution": {
56
- "iopub.execute_input": "2026-02-02T13:04:39.111866Z",
57
- "iopub.status.busy": "2026-02-02T13:04:39.111687Z",
58
- "iopub.status.idle": "2026-02-02T13:04:39.115457Z",
59
- "shell.execute_reply": "2026-02-02T13:04:39.114883Z"
56
+ "iopub.execute_input": "2026-02-03T01:49:09.168801Z",
57
+ "iopub.status.busy": "2026-02-03T01:49:09.168645Z",
58
+ "iopub.status.idle": "2026-02-03T01:49:09.172791Z",
59
+ "shell.execute_reply": "2026-02-03T01:49:09.171776Z"
60
60
  },
61
61
  "papermill": {
62
- "duration": 0.006137,
63
- "end_time": "2026-02-02T13:04:39.116176",
62
+ "duration": 0.00748,
63
+ "end_time": "2026-02-03T01:49:09.173806",
64
64
  "exception": false,
65
- "start_time": "2026-02-02T13:04:39.110039",
65
+ "start_time": "2026-02-03T01:49:09.166326",
66
66
  "status": "completed"
67
67
  },
68
68
  "tags": []
@@ -93,16 +93,16 @@
93
93
  "id": "8f7rbn1v0my",
94
94
  "metadata": {
95
95
  "execution": {
96
- "iopub.execute_input": "2026-02-02T13:04:39.118583Z",
97
- "iopub.status.busy": "2026-02-02T13:04:39.118469Z",
98
- "iopub.status.idle": "2026-02-02T13:04:39.194067Z",
99
- "shell.execute_reply": "2026-02-02T13:04:39.193202Z"
96
+ "iopub.execute_input": "2026-02-03T01:49:09.178740Z",
97
+ "iopub.status.busy": "2026-02-03T01:49:09.178613Z",
98
+ "iopub.status.idle": "2026-02-03T01:49:09.270111Z",
99
+ "shell.execute_reply": "2026-02-03T01:49:09.269548Z"
100
100
  },
101
101
  "papermill": {
102
- "duration": 0.080946,
103
- "end_time": "2026-02-02T13:04:39.197931",
102
+ "duration": 0.09754,
103
+ "end_time": "2026-02-03T01:49:09.273402",
104
104
  "exception": false,
105
- "start_time": "2026-02-02T13:04:39.116985",
105
+ "start_time": "2026-02-03T01:49:09.175862",
106
106
  "status": "completed"
107
107
  },
108
108
  "tags": []
@@ -135,14 +135,14 @@
135
135
  },
136
136
  "papermill": {
137
137
  "default_parameters": {},
138
- "duration": 1.272056,
139
- "end_time": "2026-02-02T13:04:39.465128",
138
+ "duration": 2.960898,
139
+ "end_time": "2026-02-03T01:49:10.512682",
140
140
  "environment_variables": {},
141
141
  "exception": null,
142
142
  "input_path": "/Users/Vital/python/CustomerRetention/exploration_notebooks/12_view_documentation.ipynb",
143
143
  "output_path": "/Users/Vital/python/CustomerRetention/exploration_notebooks/12_view_documentation.ipynb",
144
144
  "parameters": {},
145
- "start_time": "2026-02-02T13:04:38.193072",
145
+ "start_time": "2026-02-03T01:49:07.551784",
146
146
  "version": "2.6.0"
147
147
  }
148
148
  },
{churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/METADATA RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: churnkit
3
- Version: 0.75.0a3
3
+ Version: 0.75.1a1
4
4
  Summary: Structured ML framework for customer churn prediction -- from exploration notebooks to production pipelines, locally or on Databricks.
5
5
  Project-URL: Homepage, https://github.com/aladjov/CR
6
6
  Project-URL: Documentation, https://github.com/aladjov/CR/wiki
@@ -27,7 +27,7 @@ Requires-Dist: kaleido>=0.2.1
27
27
  Requires-Dist: matplotlib>=3.7.0
28
28
  Requires-Dist: pandas>=2.0.0
29
29
  Requires-Dist: papermill>=2.4.0
30
- Requires-Dist: plotly>=5.15.0
30
+ Requires-Dist: plotly>=6.1.1
31
31
  Requires-Dist: pyarrow>=12.0.0
32
32
  Requires-Dist: pydantic>=2.0.0
33
33
  Requires-Dist: pyyaml>=6.0.0
{churnkit-0.75.0a3.dist-info → churnkit-0.75.1a1.dist-info}/RECORD RENAMED
@@ -1,4 +1,4 @@
1
- customer_retention/__init__.py,sha256=_NUgKIWKnGZ58WdDbvwbjPPK_5z3_HpB4V5uZJRVlX8,1114
1
+ customer_retention/__init__.py,sha256=eWE6DURL9I4j4mhLjGhGbuT_Ab6c_yj0tp77DF7X6yk,1114
2
2
  customer_retention/cli.py,sha256=Wdl540cZgu_9mV-hWmTV9jD3S8QTDR8Ik-5hQXYCvmg,2466
3
3
  customer_retention/analysis/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
4
4
  customer_retention/analysis/jupyter_save_hook.py,sha256=iiNFIL83yOPX8BGUjCE6Pt5Kc8X-2adtE1_NZTMUaZQ,947
@@ -14,7 +14,7 @@ customer_retention/analysis/auto_explorer/recommendation_builder.py,sha256=7edPc
14
14
  customer_retention/analysis/auto_explorer/recommendations.py,sha256=klFBv71bubYB1Tiz9c7SqMd80hp6pp9EvQHgs-u3wDA,19552
15
15
  customer_retention/analysis/business/__init__.py,sha256=eSwOaKbkTNS9MgYiHCoI6GrAMWr7C0II-DgddVy4e1g,1404
16
16
  customer_retention/analysis/business/ab_test_designer.py,sha256=fGDz7Vd42jseMt7UFh3xwlPngDbS8K1X3LjBQd7F6l4,5911
17
- customer_retention/analysis/business/fairness_analyzer.py,sha256=nLrRP4vEC8qAq8cC6FXnpjZdY4EunR4aco9VfZyUKHg,6397
17
+ customer_retention/analysis/business/fairness_analyzer.py,sha256=3Xq-ZqAkhBJ81nhR7rxq5Wqx-BnTPtKRJT3llbBlthY,6395
18
18
  customer_retention/analysis/business/intervention_matcher.py,sha256=vHEtqW0aLsU9C_5Jc7PL7qEENGSnvImGHCXYG0CLK7k,5222
19
19
  customer_retention/analysis/business/report_generator.py,sha256=iGlDIUbrraBGs5I1IKHbiIQcojlImjd9dBeru1J-moU,9500
20
20
  customer_retention/analysis/business/risk_profile.py,sha256=tS585zNKOy31I8hZAmQ5RmuA_Df6LMRr4jLMLGnU9c0,8238
@@ -26,13 +26,13 @@ customer_retention/analysis/diagnostics/error_analyzer.py,sha256=vAvse-I5I0vXEZu
26
26
  customer_retention/analysis/diagnostics/leakage_detector.py,sha256=M3Jtp8X-V75XGZKLkfcWbBTNAgxKWxQebiZQYXnw0tY,19390
27
27
  customer_retention/analysis/diagnostics/noise_tester.py,sha256=g4B8cJpcdFjs_CUE04dEWwIp34GrVanLEgaXfj9jMaw,6512
28
28
  customer_retention/analysis/diagnostics/overfitting_analyzer.py,sha256=Ljw0Vylgfz7pUnNx7yXyJF8IBkZZ5LvSl3JspIUhmkw,8897
29
- customer_retention/analysis/diagnostics/segment_analyzer.py,sha256=DTjdCOnQufEdv52AoTRmfze6SXkgne-oOOdSjak9qZ0,5319
29
+ customer_retention/analysis/diagnostics/segment_analyzer.py,sha256=s7-DSWkXK26XRxtdgRfKTADg9FeeiX7mDmCaWY_m8uE,5324
30
30
  customer_retention/analysis/discovery/__init__.py,sha256=a3nRPb1yaZmLZaVoOqMPAqt_cc6auZisccF7hra4kBQ,342
31
31
  customer_retention/analysis/discovery/config_generator.py,sha256=mJQU3olytOuHwOF2zNKsu0a1jtt3cUxOzzNtzh1SBpA,1872
32
32
  customer_retention/analysis/discovery/discovery_flow.py,sha256=BecdqYtIINjx7mrePqJtejvXn4RYyQreXukfr-y8P_0,668
33
33
  customer_retention/analysis/discovery/type_inferencer.py,sha256=lg1yX1gkJYjViw1UpJTy4macu7i_PZ6APiqTPKIWgO4,6660
34
34
  customer_retention/analysis/interpretability/__init__.py,sha256=FCOcr4AqhlfVszSm18aqniRXkAXue0kTDKlnA6No3YE,820
35
- customer_retention/analysis/interpretability/cohort_analyzer.py,sha256=xWFTOmYqJ906r18QAM-SGgbIEHnfGiHkxMku0YoR8zE,7801
35
+ customer_retention/analysis/interpretability/cohort_analyzer.py,sha256=tylKZegWww-wSlnKtjeAUCyKS6yn6OSDIYZ_u1Wm44Q,7793
36
36
  customer_retention/analysis/interpretability/counterfactual.py,sha256=7DNkVpfK3CLpAWVtyAOLNMpUSkcD2Rvuj0Bq1WrHxiM,7755
37
37
  customer_retention/analysis/interpretability/individual_explainer.py,sha256=uhPgAe61YNSBGThSmd1cbwAtXe3nXqi9HoplmnAcMTo,5567
38
38
  customer_retention/analysis/interpretability/pdp_generator.py,sha256=G2vDQnaw_t8QTPJrVWr9pLbZ32M8Gv_jbmRebXkKv8A,3972
@@ -63,7 +63,7 @@ customer_retention/analysis/visualization/number_formatter.py,sha256=I1gUB0tEmfT
63
63
  customer_retention/artifacts/__init__.py,sha256=zTROqiS6zlkkuCZgR6YOB0Cvlsyr0TpRBYsOEorpDYw,118
64
64
  customer_retention/artifacts/fit_artifact_registry.py,sha256=aNfZC0Dgbc6jEwRR5keDEop9jo_tuL82hKO3ouCh5eY,5750
65
65
  customer_retention/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
66
- customer_retention/core/compat/__init__.py,sha256=PIz859jJ-naUGCxXkMrDAvtnZIc12AenRkqzN3UYQ1c,4930
66
+ customer_retention/core/compat/__init__.py,sha256=tD0XOABGAVQwuUaII8v1EKVJ149n1anp2dUzQkFkl-M,5382
67
67
  customer_retention/core/compat/detection.py,sha256=6W_1LefgQriBtRY2PnvSCUGDt0X63oIUEEVjFqG3qH0,2492
68
68
  customer_retention/core/compat/ops.py,sha256=L-tAh4A3UEfRvePS6rAbhqb0QtZ_bN-TV7ZWpTkMFLA,1809
69
69
  customer_retention/core/compat/pandas_backend.py,sha256=14JPoYTW6X-a3UwFaemhmPr8zi_GTdZnyitmqPQODR0,1839
@@ -180,12 +180,12 @@ customer_retention/stages/deployment/champion_challenger.py,sha256=WyiBzcQU26Ivi
180
180
  customer_retention/stages/deployment/model_registry.py,sha256=igVCM5iCNYuCdwsN6XFruAymYpluQgSJMriR6LzJQx4,7194
181
181
  customer_retention/stages/deployment/retraining_trigger.py,sha256=3ouGaWm_p9DEtopAbD6qFhsfumK-sa20bBm7eNFs__w,10172
182
182
  customer_retention/stages/features/__init__.py,sha256=vj_mkzTXZD3T6kNfkFOUU8kgL4I3ypZp8tc-MX9fHGU,1937
183
- customer_retention/stages/features/behavioral_features.py,sha256=0d_NbDuhZKqTETdn5stHzofsonzmOC45HWr1FqjEZlE,10272
184
- customer_retention/stages/features/customer_segmentation.py,sha256=LzKRG4wFW-ZcTUKqJsUekAEUCOEz3jF8Zq9NlrIf2iI,17786
183
+ customer_retention/stages/features/behavioral_features.py,sha256=aU7up9yX2tUn0QPdY3vnYosccRbJeOzthheyXTuKpNs,10280
184
+ customer_retention/stages/features/customer_segmentation.py,sha256=p41vuA3E7c4P3Bta4soePFeFRGjeyiT6DbbN-REdnYM,17779
185
185
  customer_retention/stages/features/feature_definitions.py,sha256=EvmdaxM1HZ_oWy6BfmD4MFGb4FVKegfcs4AQ5nNNWV0,7217
186
186
  customer_retention/stages/features/feature_engineer.py,sha256=btVsdLHRKYk6E5xI-9oil93-mWYABJUc8_w7kSedJnM,21634
187
187
  customer_retention/stages/features/feature_manifest.py,sha256=EEBG7kdU_jWNcnDqdLHONIaJ-n2GcqLkjXjIxo3zn9w,9731
188
- customer_retention/stages/features/feature_selector.py,sha256=nDcoFWxhROw-wqi331ool9IyXnDDxRUzBI4bPgZSQ5M,10090
188
+ customer_retention/stages/features/feature_selector.py,sha256=_CG3ZKVuZuqrwV9YvYvlTnXf0ahhsZNLmSRhf4nwNiQ,10089
189
189
  customer_retention/stages/features/interaction_features.py,sha256=P7aaHALbFcfEchJsesVPhVmOm-v2VmYkG90t8p2tNVA,4634
190
190
  customer_retention/stages/features/temporal_features.py,sha256=0RghOQwWte7cGULbKAtMggX9pQGwYGU1f4lkl-Y_5ao,9283
191
191
  customer_retention/stages/ingestion/__init__.py,sha256=kYVOe8kq7S0I_tjY-BcdZ1IsNWrYYjzDmoAcV2lhijQ,308
@@ -220,7 +220,7 @@ customer_retention/stages/profiling/pattern_analysis_config.py,sha256=RRxrZqTA_X
220
220
  customer_retention/stages/profiling/profile_result.py,sha256=NKKh1u2FmfBqnIbOEiqBh25IZDMm91h38RT7wzA8yQI,6350
221
221
  customer_retention/stages/profiling/quality_checks.py,sha256=ov8opsY4AoM9D6Yr_fGXsVwXfpmO0OeFfhdML-xfoIM,65678
222
222
  customer_retention/stages/profiling/relationship_detector.py,sha256=9WMM8YOIl-EWPY2P3PFuOENM9D1nm5lU5sDfZTE_chQ,9477
223
- customer_retention/stages/profiling/relationship_recommender.py,sha256=vPd4DyTcZkguTvqbDF4PKnIA1k5N-HfEiV5P2H83Fe8,20132
223
+ customer_retention/stages/profiling/relationship_recommender.py,sha256=WJDMCkE9Lt_Lt5R9BtNfYBmdOWLdIX4xwAMaLipn0co,20139
224
224
  customer_retention/stages/profiling/report_generator.py,sha256=XEKjwot0dKGWOMyOPgFHaZtmi7FYZnmOdm1reRLcBHg,16607
225
225
  customer_retention/stages/profiling/scd_analyzer.py,sha256=PQvGVIL3bRCSfguarsJmW4QJXJNKdaUBnRNpyn-NDL0,5334
226
226
  customer_retention/stages/profiling/segment_analyzer.py,sha256=nMRKrHs3gsgczbqTw7DB_-FjxC256RO2juR4GqXpG38,21945
@@ -228,16 +228,16 @@ customer_retention/stages/profiling/segment_aware_outlier.py,sha256=PS5GXnf_g3D9
228
228
  customer_retention/stages/profiling/target_level_analyzer.py,sha256=XPhdHqTdK9zzBDqy-JyrTi6NFf07wRwIGsVEOAiR_dE,10491
229
229
  customer_retention/stages/profiling/temporal_analyzer.py,sha256=PXf4pYNcszp7N8_14MKFKXDku-fw2M_NLWN7jUsHd1Q,16102
230
230
  customer_retention/stages/profiling/temporal_coverage.py,sha256=r23s1qyB7o11ab_TTLOgb4q29OPA_crRshFpMLt4t_w,18561
231
- customer_retention/stages/profiling/temporal_feature_analyzer.py,sha256=iPA9UE65rYPTWO45n5_SHJGaUFSDXGjXYxZdvZa9Ep0,32330
231
+ customer_retention/stages/profiling/temporal_feature_analyzer.py,sha256=iWcT84ly5iPqNRnxDxe458R4Iha5u72_g-2-ZNAk4Gs,32343
232
232
  customer_retention/stages/profiling/temporal_feature_engineer.py,sha256=Eovymy6qoFIOYy3-sQZyqVTYfIXOAf11aomBcjLjMSE,27096
233
- customer_retention/stages/profiling/temporal_pattern_analyzer.py,sha256=o-0jkdWM0VT4Ij0ASNLZM9jy8dA0KxHG3ZWKp0FxZ0w,26381
234
- customer_retention/stages/profiling/temporal_quality_checks.py,sha256=Sl74EcHfBCFgTDonvJIp2N1RKBoWNmJoEmaqjqRLaHs,13547
233
+ customer_retention/stages/profiling/temporal_pattern_analyzer.py,sha256=VXjKr21CEaHQY7PYrasGeDWESlLeQdam64JdNB46-_Y,26383
234
+ customer_retention/stages/profiling/temporal_quality_checks.py,sha256=02teyAdAsASWsJZD7SElfutBYwK23DaNSRO4rgW1Fxs,13552
235
235
  customer_retention/stages/profiling/temporal_target_analyzer.py,sha256=eeZlUhTWZfCftwgm_dySi1feRLuoU9SRLL_r_4jgN5g,8785
236
236
  customer_retention/stages/profiling/text_embedder.py,sha256=ck7WIq7pGC7xgEzMQr7fYdHcJegYR6wfdh3z32WUiK8,3038
237
237
  customer_retention/stages/profiling/text_processor.py,sha256=spdfwVSEU07aYbl2bIsg_INOBt3Js-IA15WVkjf1ask,4474
238
238
  customer_retention/stages/profiling/text_reducer.py,sha256=ilSuUAu0dHUyRGTNg8TzoCEd-EAyXKvoAm4uGqwlSQs,2409
239
- customer_retention/stages/profiling/time_series_profiler.py,sha256=sFFIwumlOT5e-j2kXiIUkUrCdE59XpsAj5gzbh7xuzg,10338
240
- customer_retention/stages/profiling/time_window_aggregator.py,sha256=8DuztlesmS-mgKRbaYuEINnkR4kd-b2_zq7PvdKKE3Q,15841
239
+ customer_retention/stages/profiling/time_series_profiler.py,sha256=9EaCRHMIrSVGFW_IWnU8BU4v44WZKMb8Tbbdw13y41Y,10384
240
+ customer_retention/stages/profiling/time_window_aggregator.py,sha256=NcxVAor8JhDnPLap_iS3z81flejd7uPeTQLwLmQn7MA,15868
241
241
  customer_retention/stages/profiling/type_detector.py,sha256=VgYHWcBGepyJKNdY1FKgb9scOaosN6fDY_-WiTjfoAg,14726
242
242
  customer_retention/stages/profiling/window_recommendation.py,sha256=Apd_PDFpo49HJJzldTcwzzgJjBzEfd8mbGboBwHhzGw,13354
243
243
  customer_retention/stages/temporal/__init__.py,sha256=f86XiSUMKQgeTLyOsu89IJcafOPjdBIR9bH_hhrY8b8,6135
@@ -253,50 +253,50 @@ customer_retention/stages/temporal/timestamp_discovery.py,sha256=b2NpXvmWg4K6SBo
253
253
  customer_retention/stages/temporal/timestamp_manager.py,sha256=EisQM4_e14wsdqVxzYXkMBS5tXKIYsTbPwgwZGl5lWU,10635
254
254
  customer_retention/stages/transformation/__init__.py,sha256=6XQGYKYNqdOuxlX6IujtVqRZ099pS8X_ATd6mLqwVtQ,783
255
255
  customer_retention/stages/transformation/binary_handler.py,sha256=ObwL90YP3ivwOJONBikzZouUoBz-YCTcxWybfwA5ddc,3201
256
- customer_retention/stages/transformation/categorical_encoder.py,sha256=_eAarKNEbKZF7o0nJ5oJ29GPQ3I8a8vFCfagFfjnUyo,10140
256
+ customer_retention/stages/transformation/categorical_encoder.py,sha256=T0mLgJ6cf2kLkha4HclAeeaxlz7cVJBWYEsEt8fs5KA,10145
257
257
  customer_retention/stages/transformation/datetime_transformer.py,sha256=iWzxb7gdpn1uEPo96_ir9hDcqCERnVPhBLTTQyxq1xk,3619
258
258
  customer_retention/stages/transformation/numeric_transformer.py,sha256=wqC2aUfXargeOph8d9F4P2wLet4lnFOKoI9x1mpJucw,6367
259
- customer_retention/stages/transformation/pipeline.py,sha256=1cSpyXwoBM1xPQj308GifjVuoqBwtPeSIQXt4ke3ttU,11278
259
+ customer_retention/stages/transformation/pipeline.py,sha256=qqbpisjN4uZ050eishlEj037u2mPKEwxGG0o7GruoQM,11278
260
260
  customer_retention/stages/validation/__init__.py,sha256=8Klgpez2ApVM1n1HUWcaGjaa21-aC-ReaZIVj7zHFh4,2380
261
261
  customer_retention/stages/validation/adversarial_scoring_validator.py,sha256=DnVFEJPMK9Wech9dXdBmFuwprNL91wdQA6inFISZQow,8033
262
262
  customer_retention/stages/validation/business_sense_gate.py,sha256=0kLVQOvwdaJn1dWUcfXI9mpX-eOh7AEsfTHyfbBRsVQ,7413
263
- customer_retention/stages/validation/data_quality_gate.py,sha256=w_q1L7RWpgyh4YyNZyU1t-OJC6fCzZpVGj1oCt4-PWQ,9230
264
- customer_retention/stages/validation/data_validators.py,sha256=JoqXoiF603IV65lHU6GvEy3O9XB30-BPmqgpbLTkq2s,18155
263
+ customer_retention/stages/validation/data_quality_gate.py,sha256=MNnMptzPrh1jlZEpQsABjGSb-k6H518dXPlFDOgOtWo,9239
264
+ customer_retention/stages/validation/data_validators.py,sha256=51QIIY1vW66fczERC5l7QPA0eBko_h75IXsVLAXj5Eg,18177
265
265
  customer_retention/stages/validation/feature_quality_gate.py,sha256=M43K9PH44LXN6H14L3RBw3ZGlJ6vd_c74w_9QF6PzjA,7902
266
266
  customer_retention/stages/validation/gates.py,sha256=jSGJMSXGpvlwVl3DBAiK7xAl0Nxg1eibnnBg-jz0Mlk,4060
267
- customer_retention/stages/validation/leakage_gate.py,sha256=gtMBAEB7HIg8CuRJeewlKv3gabJJ2lw4Ku1E0bEMNNM,14042
267
+ customer_retention/stages/validation/leakage_gate.py,sha256=dopFvV4a_JIOkffrKYrYTnT6N1QP4073DbVRJmH-bG0,14046
268
268
  customer_retention/stages/validation/model_validity_gate.py,sha256=kNI6NAdQnj9qQqQ-IimXi_1ah92peq6FIJnHVwcZpqc,8370
269
269
  customer_retention/stages/validation/pipeline_validation_runner.py,sha256=uuBvGx1ej4GpYo97lfLgQEOdU1ykG0jm46Gxu2qNKtw,11000
270
270
  customer_retention/stages/validation/quality_scorer.py,sha256=VuhSEZj3rL5URvxSjdIryOS1W0x7y_BNlX5yog4ExNk,19017
271
271
  customer_retention/stages/validation/rule_generator.py,sha256=E6jeWMeCiMRq9lhoryGB8Tvdo65poJi5dj3oNRtC19k,2139
272
272
  customer_retention/stages/validation/scoring_pipeline_validator.py,sha256=FvM7AsNpiOO0nLr6NGkJGzJfUhxvPNZ7ccDrp44zFiI,21537
273
- customer_retention/stages/validation/timeseries_detector.py,sha256=ggnBmIZN7pi89LOk6iVXINoTLpZQ-AI1-pddjCokyy4,27606
273
+ customer_retention/stages/validation/timeseries_detector.py,sha256=3IJ6H9bdu3nvxJZcrbhRJpl0nY7wLo2SAukM3AlgIe4,27619
274
274
  customer_retention/transforms/__init__.py,sha256=W9owOGVCFSoCQfcRAQciNASYqbPpDE9gGjvnLcXawrE,1320
275
275
  customer_retention/transforms/artifact_store.py,sha256=FYLpDcv2N6-dUTX5RPEIK3aCWKhYK3hRpPROidLpRik,1641
276
276
  customer_retention/transforms/executor.py,sha256=oML5dCidxbW_q6YUkAwWcutYP6bIFB6IdD3BvemK45A,6304
277
277
  customer_retention/transforms/fitted.py,sha256=3pNvnae-P3t3bKMeZz1Bl0xww-feapIYdoeTY6aUtI8,3278
278
- customer_retention/transforms/ops.py,sha256=URkQcFyb9tYzteuIFR-pOyT47vnFqx95EciaQtKrdjU,4284
279
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/00_start_here.ipynb,sha256=GvWqC8sZROCKdue2SuEEenZE36zTT6Qo5Ntn4oL66mw,21909
280
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01_data_discovery.ipynb,sha256=DRZqtXvNu2KylseGfX0fuCuz5sFxLhRkvdnGH5TxT9o,46434
281
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01a_a_temporal_text_deep_dive.ipynb,sha256=uai8T3iJSqOrabBQnVi8Z0k8zZGVgs_VVQWRHyXN8QU,33690
282
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01a_temporal_deep_dive.ipynb,sha256=8s_JK6Dq1VE6mwUHJH2B3kizLZVD0iMQFgYtMArfHuk,67322
283
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01b_temporal_quality.ipynb,sha256=xmbgKfWk0T0VcpfyxafsZfQwJV-nGlXH7KFsCCHg4ns,23037
284
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01c_temporal_patterns.ipynb,sha256=Jg9MxezUrmvYtV9Jg0M1IbX7QWeRZW5l8FDv0hlh2sM,153163
285
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/01d_event_aggregation.ipynb,sha256=E7pBamAW_ChoUszssVEuNsWldn4ap40niJTzQmn-amk,63243
286
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/02_column_deep_dive.ipynb,sha256=dEbN2tR7bMArgj_lY56JY5cKLCFI4MpY_4Tv7_NwSO8,60251
287
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/02a_text_columns_deep_dive.ipynb,sha256=M9YN8yAjjuC6ZaUlc-rVqVLEkWd7Rc_GNILHS9qO3PU,29704
288
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/03_quality_assessment.ipynb,sha256=JJOCZeugEqGZbVulXF3W07ftxIoXmJb2r1U3yFRJMTg,68600
289
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/04_relationship_analysis.ipynb,sha256=cKs0J_Yjb7BKTAC_gnxFenrPPD_aUmYSCiZA10-hSxM,78622
290
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/05_multi_dataset.ipynb,sha256=HveFZnqIfBb0dY6IbXZDa1Bj2Xat1067Ip0bsJLoA7M,62667
291
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/06_feature_opportunities.ipynb,sha256=O4XIgV1SpNYpvi_08qtqwCd05XdGlA-ZaOEl6tI7eeA,67409
292
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/07_modeling_readiness.ipynb,sha256=KNRtShdXyJnMnDlBIAnOoLvzzdMzc_TiQ_ug0J7n6kI,28376
293
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/08_baseline_experiments.ipynb,sha256=YXYZLtG9wul69Nj6IOk39yfQ4SRhL502Zn5ynHVdotA,34228
294
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/09_business_alignment.ipynb,sha256=sAtCIaFQ1IHWuvzc55WmY7SGsPqZntE8HHn5Q0qcTpA,17216
295
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/10_spec_generation.ipynb,sha256=g_ott0EhnVVPltgTvJeSboavY-4IJaLKvzs71AMbTtw,42843
296
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/11_scoring_validation.ipynb,sha256=z5xpgh0BHJQ2-3cQk6PwlQ7roGpMvXyJZr8v_YfAsW8,49938
297
- churnkit-0.75.0a3.data/data/share/churnkit/exploration_notebooks/12_view_documentation.ipynb,sha256=kIjL-z1mt3cshgawIROZLPMsQBBEhxOYUoXXDNARR8w,4407
298
- churnkit-0.75.0a3.dist-info/METADATA,sha256=lQIZJ7CLPcMAMfu0Jjj8ZsN20QvgTK4EuIIqAsOAFL8,12737
299
- churnkit-0.75.0a3.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
300
- churnkit-0.75.0a3.dist-info/entry_points.txt,sha256=swQFVe-jjgQSBJQNO2Ulkz2F5odaE-TsnlTor3HQBjw,70
301
- churnkit-0.75.0a3.dist-info/licenses/LICENSE,sha256=Bud8Oj25tnpoIuXCWW0xcSfmGPeEZAAHrDRoKdSYtZY,11344
302
- churnkit-0.75.0a3.dist-info/RECORD,,
278
+ customer_retention/transforms/ops.py,sha256=Xg2g9UOOudq_y9Hf3oWsjpqw3dEoykQR5pDSoyW8GX0,4294
279
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/00_start_here.ipynb,sha256=zCyhftEd3v9fc0Ta6wvA6b-9LcoGzRi8bS1tMZ3iu9w,21911
280
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01_data_discovery.ipynb,sha256=up0X3oDJ5sAo1-tbqMyZj_f1h6D542G2uAxjVmtYCOI,46430
281
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01a_a_temporal_text_deep_dive.ipynb,sha256=uai8T3iJSqOrabBQnVi8Z0k8zZGVgs_VVQWRHyXN8QU,33690
282
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01a_temporal_deep_dive.ipynb,sha256=fC1ASNtvI8X1lAe-Lzcw3oX2cptDC-ymPeEtKKWhg20,67326
283
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01b_temporal_quality.ipynb,sha256=RU5hxgrTVMZs1ytChVv1t49WpTO0Oj6B_Fu8g0xS0To,23039
284
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01c_temporal_patterns.ipynb,sha256=ZGYfztP6JhOEwPmTYdC0l7w579fKXcNEJXq-PnCLc2I,153167
285
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/01d_event_aggregation.ipynb,sha256=-FT3SoBU0fhaZxGeTo-_UQl6riCrtoJaFnUg31opk64,63244
286
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/02_column_deep_dive.ipynb,sha256=mbP2LQWsXDyTsWg0bhrCBHEfHsEer_XOXRYV9f8JxAk,60250
287
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/02a_text_columns_deep_dive.ipynb,sha256=M9YN8yAjjuC6ZaUlc-rVqVLEkWd7Rc_GNILHS9qO3PU,29704
288
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/03_quality_assessment.ipynb,sha256=H49LLmn1PHbcbAvSQfteESRGk125QwkPI5qbLk3yZgc,68595
289
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/04_relationship_analysis.ipynb,sha256=Rr-B4-xg0ILuAIgztlZkiGJdTzLuNjOqBFxO8W4o9iU,78624
290
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/05_multi_dataset.ipynb,sha256=bBxkuZyTl1yZg4kMXO87WRjgZMhj_6hwLGX6m3XC270,62664
291
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/06_feature_opportunities.ipynb,sha256=cBJF5o4z3Z-dustQ4CVklnfTcQ8saG97tlgswWK9uWE,67409
292
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/07_modeling_readiness.ipynb,sha256=IiA04fyb-l097Glp3MtR03vPjQsZlS1Icg-hjEHa_Dg,28376
293
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/08_baseline_experiments.ipynb,sha256=KmjhnDf1JdpEiIcdfQ-ZFo_at6t9JRC30B6NmmvMBmg,34226
294
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/09_business_alignment.ipynb,sha256=tMNfGM7AH50N1ONzHhGW2HZLpQwraIxVzOiVnI-10X8,17214
295
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/10_spec_generation.ipynb,sha256=KeUdfL9Mvdi6023XpnfZ6oLEDNZaWiIHUfsAWig24mE,42847
296
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/11_scoring_validation.ipynb,sha256=5fi3eHMm03ZKZgdFAXMgydtZ3qX2TtR3L9bZS2MpWPE,49937
297
+ churnkit-0.75.1a1.data/data/share/churnkit/exploration_notebooks/12_view_documentation.ipynb,sha256=aQF7CG8HxckqUKOKqnmZgMkSvfVzyO2LlYPrymLYjBY,4405
298
+ churnkit-0.75.1a1.dist-info/METADATA,sha256=7IQb94O6dIEFTy7hejcQJU8eXJ6QKnnmxpyZqUKSrhY,12736
299
+ churnkit-0.75.1a1.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
300
+ churnkit-0.75.1a1.dist-info/entry_points.txt,sha256=swQFVe-jjgQSBJQNO2Ulkz2F5odaE-TsnlTor3HQBjw,70
301
+ churnkit-0.75.1a1.dist-info/licenses/LICENSE,sha256=Bud8Oj25tnpoIuXCWW0xcSfmGPeEZAAHrDRoKdSYtZY,11344
302
+ churnkit-0.75.1a1.dist-info/RECORD,,
customer_retention/__init__.py CHANGED
@@ -17,7 +17,7 @@ Main module categories:
17
17
  llm_context, iteration)
18
18
  """
19
19
 
20
- __version__ = "0.75.0a3"
20
+ __version__ = "0.75.1a1"
21
21
 
22
22
  # Environment utilities (always available)
23
23
  from .core.compat import (
customer_retention/analysis/business/fairness_analyzer.py CHANGED
@@ -3,7 +3,7 @@
3
3
  from dataclasses import dataclass
4
4
  from typing import Dict, List, Optional
5
5
 
6
- from customer_retention.core.compat import Series, pd
6
+ from customer_retention.core.compat import Series, cut
7
7
 
8
8
 
9
9
  @dataclass
@@ -138,7 +138,7 @@ class FairnessAnalyzer:
138
138
  mask = protected == group
139
139
  y_t = y_true[mask]
140
140
  y_p = y_proba[mask]
141
- bins = pd.cut(y_p, bins=10, labels=False)
141
+ bins = cut(y_p, bins=10, labels=False)
142
142
  calibration_error = 0
143
143
  for b in range(10):
144
144
  bin_mask = bins == b
customer_retention/analysis/diagnostics/segment_analyzer.py CHANGED
@@ -6,7 +6,7 @@ from typing import Dict, List
6
6
  import numpy as np
7
7
  from sklearn.metrics import average_precision_score, precision_score, recall_score, roc_auc_score
8
8
 
9
- from customer_retention.core.compat import DataFrame, Series, pd
9
+ from customer_retention.core.compat import DataFrame, Series, cut, pd, qcut
10
10
  from customer_retention.core.components.enums import Severity
11
11
 
12
12
 
@@ -39,9 +39,9 @@ class SegmentPerformanceAnalyzer:
39
39
  return pd.Series(["all"] * len(X))
40
40
  values = X[segment_column]
41
41
  if segment_type == "tenure":
42
- return pd.cut(values, bins=[0, 90, 365, np.inf], labels=["new", "established", "mature"])
42
+ return cut(values, bins=[0, 90, 365, np.inf], labels=["new", "established", "mature"])
43
43
  if segment_type == "quantile":
44
- return pd.qcut(values, q=3, labels=["low", "medium", "high"], duplicates="drop")
44
+ return qcut(values, q=3, labels=["low", "medium", "high"], duplicates="drop")
45
45
  return Series(["all"] * len(X))
46
46
 
47
47
  def analyze_performance(self, model, X: DataFrame, y: Series, segments: Series) -> SegmentResult:
customer_retention/analysis/interpretability/cohort_analyzer.py CHANGED
@@ -6,7 +6,7 @@ from typing import Any, Dict, List
6
6
  import numpy as np
7
7
  import shap
8
8
 
9
- from customer_retention.core.compat import DataFrame, Series, pd
9
+ from customer_retention.core.compat import DataFrame, Series, cut
10
10
 
11
11
 
12
12
  @dataclass
@@ -167,19 +167,19 @@ class CohortAnalyzer:
167
167
  bins: List[float] = None) -> Series:
168
168
  bins = bins or [0, 90, 365, float("inf")]
169
169
  labels = ["New", "Established", "Mature"]
170
- return pd.cut(tenure, bins=bins, labels=labels)
170
+ return cut(tenure, bins=bins, labels=labels)
171
171
 
172
172
  @staticmethod
173
173
  def create_value_cohorts(value: Series,
174
174
  quantiles: List[float] = None) -> Series:
175
175
  quantiles = quantiles or [0.33, 0.66]
176
176
  q1, q2 = value.quantile(quantiles[0]), value.quantile(quantiles[1])
177
- return pd.cut(value, bins=[-float("inf"), q1, q2, float("inf")],
177
+ return cut(value, bins=[-float("inf"), q1, q2, float("inf")],
178
178
  labels=["Low", "Medium", "High"])
179
179
 
180
180
  @staticmethod
181
181
  def create_activity_cohorts(activity: Series,
182
182
  thresholds: List[float] = None) -> Series:
183
183
  thresholds = thresholds or [5, 15]
184
- return pd.cut(activity, bins=[-float("inf"), thresholds[0], thresholds[1], float("inf")],
184
+ return cut(activity, bins=[-float("inf"), thresholds[0], thresholds[1], float("inf")],
185
185
  labels=["Dormant", "Moderate", "Active"])
customer_retention/core/compat/__init__.py CHANGED
@@ -105,6 +105,17 @@ CategoricalDtype = _pandas.CategoricalDtype
105
105
  NA = _pandas.NA
106
106
  NaT = _pandas.NaT
107
107
 
108
+ # Utility functions (always use real pandas, never pyspark.pandas)
109
+ to_datetime = _pandas.to_datetime
110
+ to_timedelta = _pandas.to_timedelta
111
+ to_numeric = _pandas.to_numeric
112
+ cut = _pandas.cut
113
+ qcut = _pandas.qcut
114
+ get_dummies = _pandas.get_dummies
115
+ crosstab = _pandas.crosstab
116
+ notna = _pandas.notna
117
+ isna = _pandas.isna
118
+
108
119
  api_types = _pandas.api.types
109
120
 
110
121
 
@@ -162,6 +173,15 @@ __all__ = [
162
173
  "CategoricalDtype",
163
174
  "NA",
164
175
  "NaT",
176
+ "to_datetime",
177
+ "to_timedelta",
178
+ "to_numeric",
179
+ "cut",
180
+ "qcut",
181
+ "get_dummies",
182
+ "crosstab",
183
+ "notna",
184
+ "isna",
165
185
  "is_spark_available",
166
186
  "is_pandas_api_on_spark",
167
187
  "get_pandas",
customer_retention/stages/features/behavioral_features.py CHANGED
@@ -10,7 +10,7 @@ from typing import List, Optional
10
10
 
11
11
  import numpy as np
12
12
 
13
- from customer_retention.core.compat import DataFrame, pd
13
+ from customer_retention.core.compat import DataFrame, cut, to_datetime
14
14
 
15
15
 
16
16
  @dataclass
@@ -238,7 +238,7 @@ class BehavioralFeatureGenerator:
238
238
  """Generate recency bucket feature."""
239
239
  if self.days_since_last_order_column:
240
240
  if self.days_since_last_order_column in df.columns:
241
- df["recency_bucket"] = pd.cut(
241
+ df["recency_bucket"] = cut(
242
242
  df[self.days_since_last_order_column],
243
243
  bins=self.recency_bins,
244
244
  labels=self.recency_labels,
@@ -253,7 +253,7 @@ class BehavioralFeatureGenerator:
253
253
  if self.feature_timestamp_column not in df.columns:
254
254
  return
255
255
 
256
- feature_ts = pd.to_datetime(df[self.feature_timestamp_column], format='mixed')
256
+ feature_ts = to_datetime(df[self.feature_timestamp_column], format='mixed')
257
257
  datetime_cols = df.select_dtypes(include=["datetime64"]).columns
258
258
 
259
259
  for col in datetime_cols:
customer_retention/stages/features/customer_segmentation.py CHANGED
@@ -9,7 +9,7 @@ from dataclasses import dataclass, field
9
9
  from enum import Enum
10
10
  from typing import Any, Dict, List, Optional
11
11
 
12
- from customer_retention.core.compat import DataFrame, pd
12
+ from customer_retention.core.compat import DataFrame, is_datetime64_any_dtype, isna, to_datetime
13
13
 
14
14
 
15
15
  class SegmentationType(Enum):
@@ -220,7 +220,7 @@ class CustomerSegmenter:
220
220
  lapsing_days = thresholds.get("lapsing", 180)
221
221
 
222
222
  def assign_recency_bucket(days):
223
- if pd.isna(days):
223
+ if isna(days):
224
224
  return "Unknown"
225
225
  days = int(days)
226
226
  if days <= active_days:
@@ -300,7 +300,7 @@ class CustomerSegmenter:
300
300
  df_result = df.copy()
301
301
 
302
302
  def assign_engagement(score):
303
- if pd.isna(score):
303
+ if isna(score):
304
304
  return "Unknown"
305
305
  if score >= high_threshold:
306
306
  return "High_Engagement"
@@ -428,14 +428,14 @@ class CustomerSegmenter:
428
428
  df_result = df.copy()
429
429
 
430
430
  # Ensure datetime
431
- if not pd.api.types.is_datetime64_any_dtype(df_result[created_column]):
432
- df_result[created_column] = pd.to_datetime(df_result[created_column], errors='coerce', format='mixed')
431
+ if not is_datetime64_any_dtype(df_result[created_column]):
432
+ df_result[created_column] = to_datetime(df_result[created_column], errors='coerce', format='mixed')
433
433
 
434
434
  # Set reference date
435
435
  if reference_date is None:
436
436
  reference_date = df_result[created_column].max()
437
437
  else:
438
- reference_date = pd.to_datetime(reference_date)
438
+ reference_date = to_datetime(reference_date)
439
439
 
440
440
  prefix = f"{output_prefix}_" if output_prefix else ""
441
441
 
@@ -447,7 +447,7 @@ class CustomerSegmenter:
447
447
 
448
448
  # Tenure bucket
449
449
  def tenure_bucket(days):
450
- if pd.isna(days) or days < 0:
450
+ if isna(days) or days < 0:
451
451
  return "Unknown"
452
452
  if days <= 90:
453
453
  return "New_0_3m"
@@ -491,14 +491,14 @@ class CustomerSegmenter:
491
491
  df_result = df.copy()
492
492
 
493
493
  # Ensure datetime
494
- if not pd.api.types.is_datetime64_any_dtype(df_result[last_activity_column]):
495
- df_result[last_activity_column] = pd.to_datetime(df_result[last_activity_column], errors='coerce', format='mixed')
494
+ if not is_datetime64_any_dtype(df_result[last_activity_column]):
495
+ df_result[last_activity_column] = to_datetime(df_result[last_activity_column], errors='coerce', format='mixed')
496
496
 
497
497
  # Set reference date
498
498
  if reference_date is None:
499
499
  reference_date = df_result[last_activity_column].max()
500
500
  else:
501
- reference_date = pd.to_datetime(reference_date)
501
+ reference_date = to_datetime(reference_date)
502
502
 
503
503
  df_result[output_column] = (reference_date - df_result[last_activity_column]).dt.days
504
504
 
customer_retention/stages/features/feature_selector.py CHANGED
@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING, Any, Dict, List, Optional
4
4
 
5
5
  import numpy as np
6
6
 
7
- from customer_retention.core.compat import DataFrame, is_numeric_dtype, pd
7
+ from customer_retention.core.compat import DataFrame, is_numeric_dtype, isna
8
8
 
9
9
  if TYPE_CHECKING:
10
10
  from customer_retention.analysis.auto_explorer.findings import FeatureAvailabilityMetadata
@@ -126,7 +126,7 @@ class FeatureSelector:
126
126
  continue
127
127
 
128
128
  variance = series.var()
129
- if pd.isna(variance) or variance < self.variance_threshold:
129
+ if isna(variance) or variance < self.variance_threshold:
130
130
  if feature in self.selected_features:
131
131
  self.selected_features.remove(feature)
132
132
  self.dropped_features.append(feature)
customer_retention/stages/profiling/relationship_recommender.py CHANGED
@@ -6,7 +6,7 @@ from typing import Any, Dict, List, Optional
6
6
 
7
7
  import numpy as np
8
8
 
9
- from customer_retention.core.compat import pd
9
+ from customer_retention.core.compat import crosstab, pd
10
10
 
11
11
 
12
12
  class RecommendationCategory(Enum):
@@ -323,7 +323,7 @@ class RelationshipRecommender:
323
323
  """Calculate Cramér's V for categorical association."""
324
324
  try:
325
325
  from scipy.stats import chi2_contingency
326
- contingency = pd.crosstab(df[col], df[target_col])
326
+ contingency = crosstab(df[col], df[target_col])
327
327
  chi2, _, _, _ = chi2_contingency(contingency)
328
328
  n = len(df)
329
329
  min_dim = min(contingency.shape) - 1
customer_retention/stages/profiling/temporal_feature_analyzer.py CHANGED
@@ -5,7 +5,7 @@ from typing import Any, Dict, List, Optional, Tuple
5
5
  import numpy as np
6
6
  from scipy import stats
7
7
 
8
- from customer_retention.core.compat import DataFrame, pd
8
+ from customer_retention.core.compat import DataFrame, pd, qcut, to_datetime
9
9
  from customer_retention.core.utils import compute_effect_size
10
10
 
11
11
 
@@ -627,7 +627,7 @@ class TemporalFeatureAnalyzer:
627
627
 
628
628
  def _prepare_dataframe(self, df: DataFrame) -> DataFrame:
629
629
  df = df.copy()
630
- df[self.time_column] = pd.to_datetime(df[self.time_column])
630
+ df[self.time_column] = to_datetime(df[self.time_column])
631
631
  return df
632
632
 
633
633
  def _validate_event_level_target_usage(self, df: DataFrame, target_column: Optional[str]) -> None:
@@ -646,7 +646,7 @@ class TemporalFeatureAnalyzer:
646
646
  if len(df_iv) < bins * 2:
647
647
  return 0.0
648
648
  try:
649
- df_iv["bin"] = pd.qcut(df_iv["feature"], q=bins, duplicates="drop")
649
+ df_iv["bin"] = qcut(df_iv["feature"], q=bins, duplicates="drop")
650
650
  except ValueError:
651
651
  return 0.0
652
652