tracepipe 0.3.2__tar.gz → 0.3.4__tar.gz

{tracepipe-0.3.2 → tracepipe-0.3.4}/CHANGELOG.md

@@ -5,6 +5,31 @@ All notable changes to TracePipe will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## 0.3.4 - 2026-02-03
+
+### Fixed
+- **Event deduplication**: Identical events from parallel pipelines are now deduplicated
+  - When multiple DataFrames share row IDs (e.g., from `df.copy()`), same changes are recorded once
+  - Events deduplicated by `(col, old_val, new_val, operation)` signature
+  - Prevents "4 events" when only 1 logical change occurred
+
+### Added
+- `_stable_repr()` helper for robust value comparison in deduplication
+- Tests for cross-pipeline event deduplication behavior
+
+## 0.3.3 - 2026-02-03
+
+### Fixed
+- **Double-logging bug**: `df['col'] = df['col'].fillna()` now logs exactly one event, not two
+  - Fixed duplicate capture from both `_wrap_setitem` and `wrap_series_assignment`
+- **Merge warning scoping**: `tp.check(df)` now only shows warnings for merges in df's lineage
+  - Previously showed warnings from ALL merges in the session (cross-contamination)
+  - Now filters by tracking which merge steps produced the queried DataFrame's rows
+
+### Added
+- `_get_merge_stats_for_df()` helper to scope merge warnings to df's lineage
+- Tests for double-logging prevention and merge warning scoping
+
 ## 0.3.2 - 2026-02-03
 
 ### Fixed

{tracepipe-0.3.2 → tracepipe-0.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tracepipe
-Version: 0.3.2
+Version: 0.3.4
 Summary: Row-level data lineage tracking for pandas pipelines
 Project-URL: Homepage, https://github.com/tracepipe/tracepipe
 Project-URL: Documentation, https://tracepipe.github.io/tracepipe/

{tracepipe-0.3.2 → tracepipe-0.3.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "tracepipe"
-version = "0.3.2"
+version = "0.3.4"
 description = "Row-level data lineage tracking for pandas pipelines"
 readme = "README.md"
 license = {file = "LICENSE"}

{tracepipe-0.3.2 → tracepipe-0.3.4}/tests/test_lineage_through_merge.py

@@ -318,6 +318,130 @@ class TestMergeWarningLabeling:
         )
 
 
+class TestNoDoubleLogging:
+    """Tests to verify no duplicate events are logged."""
+
+    def test_fillna_logs_once_not_twice(self):
+        """df['col'] = df['col'].fillna() should log exactly one event per row."""
+        tp.enable(mode="debug", watch=["income"])
+
+        df = pd.DataFrame({"id": [1], "income": [None]})
+        df["income"] = df["income"].fillna(0)
+
+        result = tp.why(df, col="income", row=0)
+
+        # Should have exactly 1 change event, not 2 (no double-logging)
+        assert result.n_changes == 1, (
+            f"Expected exactly 1 change event, got {result.n_changes}. "
+            f"Double-logging bug if > 1. History: {result.history}"
+        )
+
+    def test_setitem_logs_once(self):
+        """Direct column assignment should log exactly once."""
+        tp.enable(mode="debug", watch=["val"])
+
+        df = pd.DataFrame({"id": [1], "val": [10]})
+        df["val"] = df["val"] * 2
+
+        result = tp.why(df, col="val", row=0)
+
+        assert result.n_changes == 1, (
+            f"Expected exactly 1 change event, got {result.n_changes}. "
+            f"History: {result.history}"
+        )
+
+    def test_cross_pipeline_identical_change_deduplication(self):
+        """Identical changes from parallel pipelines should be deduplicated.
+
+        When multiple pipelines from the same source do the SAME transformation
+        (e.g., both do fillna(0)), the event should only appear once.
+        """
+        tp.enable(mode="debug", watch=["income"])
+
+        # Source data with a row that has None income
+        customers = pd.DataFrame({"id": ["A", "B"], "income": [None, 100.0]})
+
+        # Two parallel pipelines doing the SAME transformation
+        df1 = customers.copy()
+        df1["income"] = df1["income"].fillna(0)  # Records None -> 0
+
+        df2 = customers.copy()
+        df2["income"] = df2["income"].fillna(0)  # Records SAME None -> 0
+
+        # Query df1 - should deduplicate identical events
+        result1 = tp.why(df1, col="income", row=0)
+        assert result1.n_changes == 1, (
+            f"Identical events should be deduplicated. Got {result1.n_changes}. "
+            f"History: {result1.history}"
+        )
+
+    def test_cross_pipeline_different_changes_preserved(self):
+        """Different changes from parallel pipelines should NOT be deduplicated.
+
+        When pipelines do DIFFERENT transformations on the same row,
+        both events should be visible (this is expected behavior - row IDs
+        are shared, so history includes all changes to that row ID).
+        """
+        tp.enable(mode="debug", watch=["income"])
+
+        customers = pd.DataFrame({"id": ["A", "B"], "income": [None, 100.0]})
+
+        # Two pipelines doing DIFFERENT transformations
+        df1 = customers.copy()
+        df1["income"] = df1["income"].fillna(0)  # None -> 0
+
+        df2 = customers.copy()
+        df2["income"] = df2["income"].fillna(99)  # None -> 99
+
+        # Query df1 - since row ID is shared, both changes are visible
+        # This is expected: deduplication only removes IDENTICAL events
+        result1 = tp.why(df1, col="income", row=0)
+        assert result1.n_changes == 2, (
+            f"Different changes should both be visible. Got {result1.n_changes}. "
+            f"History: {result1.history}"
+        )
+
+
+class TestMergeWarningScoping:
+    """Tests for merge warnings being scoped to df's lineage."""
+
+    def test_check_only_shows_warnings_for_df_lineage(self):
+        """check(df) should only show warnings from merges that produced df."""
+        tp.enable(mode="debug")
+
+        # First pipeline - merge with unique right
+        left1 = pd.DataFrame({"k": ["a", "a", "b"], "v": [1, 2, 3]})  # Dup left
+        right1 = pd.DataFrame({"k": ["a", "b", "c"], "r": [10, 20, 30]})  # Unique
+        df1 = left1.merge(right1, on="k", how="left")
+
+        # Second pipeline - merge with duplicate right
+        left2 = pd.DataFrame({"k": ["x", "y"], "v": [1, 2]})  # Unique
+        right2 = pd.DataFrame({"k": ["x", "x", "y"], "r": [10, 11, 20]})  # Dup right
+        df2 = left2.merge(right2, on="k", how="left")
+
+        # check(df1) should only show warnings about LEFT duplicates (from df1's merge)
+        result1 = tp.check(df1)
+        dup_warnings1 = [w for w in result1.warnings if "duplicate" in w.message.lower()]
+
+        # check(df2) should only show warnings about RIGHT duplicates (from df2's merge)
+        result2 = tp.check(df2)
+        dup_warnings2 = [w for w in result2.warnings if "duplicate" in w.message.lower()]
+
+        # df1's check should NOT include df2's "Right table" warning
+        right_in_df1 = [w for w in dup_warnings1 if "right" in w.message.lower()]
+        assert len(right_in_df1) == 0, (
+            f"df1 check should not have Right table warnings (from df2). "
+            f"Got: {[w.message for w in dup_warnings1]}"
+        )
+
+        # df2's check should NOT include df1's "Left table" warning
+        left_in_df2 = [w for w in dup_warnings2 if "left" in w.message.lower()]
+        assert len(left_in_df2) == 0, (
+            f"df2 check should not have Left table warnings (from df1). "
+            f"Got: {[w.message for w in dup_warnings2]}"
+        )
+
+
 class TestLineageDepthLimit:
     """Tests for lineage traversal depth limiting."""
 

{tracepipe-0.3.2 → tracepipe-0.3.4}/tracepipe/__init__.py

@@ -81,7 +81,7 @@ from .core import TracePipeConfig, TracePipeMode
 from .snapshot import DiffResult, Snapshot, diff, snapshot
 
 # === VERSION ===
-__version__ = "0.3.2"
+__version__ = "0.3.4"
 
 # === MINIMAL __all__ ===
 __all__ = [

{tracepipe-0.3.2 → tracepipe-0.3.4}/tracepipe/convenience.py

@@ -361,8 +361,8 @@ def check(
     facts["rows_dropped"] = len(dropped)
     facts["total_steps"] = len(ctx.store.steps)
 
-    # Merge statistics
-    merge_stats_list = ctx.store.get_merge_stats() if hasattr(ctx.store, "get_merge_stats") else []
+    # Merge statistics - filter to df's lineage to avoid cross-contamination
+    merge_stats_list = _get_merge_stats_for_df(df, ctx)
 
     for i, (step_id, stats) in enumerate(merge_stats_list):
         facts[f"merge_{i}_expansion"] = stats.expansion_ratio
@@ -658,6 +658,53 @@ def find(
 # ============ HELPERS ============
 
 
+def _get_merge_stats_for_df(df: pd.DataFrame, ctx) -> list[tuple[int, Any]]:
+    """
+    Get merge stats relevant to df's lineage only.
+
+    This prevents cross-contamination where check(df) would show warnings
+    from merges that produced OTHER DataFrames in the same session.
+    """
+    if not hasattr(ctx.store, "get_merge_stats"):
+        return []
+
+    all_stats = ctx.store.get_merge_stats()
+    if not all_stats:
+        return []
+
+    # Get row IDs from df
+    rids = ctx.row_manager.get_ids_array(df)
+    if rids is None:
+        return []
+
+    # Find which merge steps produced rows in df
+    relevant_step_ids = set()
+
+    # Check merge mappings to find which merges produced df's rows
+    if hasattr(ctx.store, "merge_mappings"):
+        for mapping in ctx.store.merge_mappings:
+            # Check if any of df's row IDs are in this merge's output
+            for rid in rids:
+                # Binary search in sorted out_rids
+                i = np.searchsorted(mapping.out_rids, rid)
+                if i < len(mapping.out_rids) and mapping.out_rids[i] == rid:
+                    relevant_step_ids.add(mapping.step_id)
+                    break  # Found at least one match, this merge is relevant
+
+    # If no merge mappings found, fall back to checking if df was just merged
+    # by seeing if it has more columns than typical (heuristic)
+    if not relevant_step_ids and all_stats:
+        # Fallback: return only the most recent merge that could have produced df
+        # This handles the case where merge_mappings aren't available
+        for step_id, stats in reversed(all_stats):
+            if stats.result_rows == len(df):
+                relevant_step_ids.add(step_id)
+                break
+
+    # Filter stats to relevant merges only
+    return [(sid, stats) for sid, stats in all_stats if sid in relevant_step_ids]
+
+
 def _json_safe(val: Any) -> Any:
     """Convert value to JSON-serializable form."""
     if pd.isna(val):

{tracepipe-0.3.2 → tracepipe-0.3.4}/tracepipe/instrumentation/series_capture.py

@@ -116,6 +116,10 @@ def wrap_series_assignment():
     """
     Wrap DataFrame.__setitem__ to capture diffs when assigning Series.
 
+    Note: For watched columns, _wrap_setitem (pandas_inst.py) already captures
+    the assignment. This wrapper only captures for NON-watched columns when
+    a TrackedSeries is assigned, to avoid double-logging.
+
     Handles:
     - df['col'] = series  (where series may have been modified)
     - df['col'] = scalar  (broadcast assignment)
@@ -127,28 +131,33 @@ def wrap_series_assignment():
     def tracked_setitem(self, key, value):
         ctx = get_context()
 
-        # Capture before state for watched columns
+        # For watched columns, _wrap_setitem already captures - skip to avoid double-logging
+        # We only capture here for NON-watched columns when a TrackedSeries is involved
+        should_capture_here = False
         before_values = None
+
         if (
             ctx.enabled
             and isinstance(key, str)
-            and key in ctx.watched_columns
             and key in self.columns
+            and key not in ctx.watched_columns  # Only capture NON-watched columns here
+            and isinstance(value, TrackedSeries)  # Only for TrackedSeries assignments
         ):
             rids = ctx.row_manager.get_ids_array(self)
             if rids is not None:
+                should_capture_here = True
                 before_values = {
                     "rids": rids.copy(),
                     "values": self[key].values.copy(),
                 }
 
-        # Always run original
+        # Always run original (which may be _wrap_setitem's wrapper)
         original_setitem(self, key, value)
 
         if not ctx.enabled:
             return
 
-        if before_values is None:
+        if not should_capture_here or before_values is None:
             return
 
         try:

{tracepipe-0.3.2 → tracepipe-0.3.4}/tracepipe/storage/lineage_store.py

@@ -32,6 +32,22 @@ from ..core import (
 from ..utils.value_capture import capture_typed_value
 
 
+def _stable_repr(val) -> str:
+    """Create a stable string representation for deduplication.
+
+    Handles NaN, None, and other values that don't compare equal to themselves.
+    """
+    if val is None:
+        return "None"
+    # Handle NaN (which doesn't equal itself)
+    try:
+        if isinstance(val, float) and val != val:  # NaN check
+            return "NaN"
+    except (TypeError, ValueError):
+        pass
+    return repr(val)
+
+
 class InMemoryLineageStore:
     """
     Columnar storage for lineage data using Structure of Arrays (SoA).
@@ -556,12 +572,15 @@ class InMemoryLineageStore:
         Follows merge lineage recursively to build complete cell provenance.
         This is essential for tracking changes that happened before merge operations.
 
+        Deduplicates events by (col, old_val, new_val, operation) signature to prevent
+        cross-pipeline contamination when multiple DataFrames share row IDs.
+
         Args:
             row_id: Row ID to trace
             max_depth: Maximum merge depth to follow (prevents infinite loops)
 
         Returns:
-            List of events in chronological order, including parent row events.
+            List of UNIQUE events in chronological order, including parent row events.
         """
         visited: set[int] = set()
 
@@ -589,7 +608,23 @@ class InMemoryLineageStore:
         # Sort by step_id to ensure chronological order across lineage
         all_events.sort(key=lambda e: e["step_id"])
 
-        return all_events
+        # Deduplicate by (col, old_val, new_val, operation) signature
+        # This prevents cross-pipeline contamination when multiple DataFrames
+        # share the same row IDs (e.g., df.copy() followed by parallel transforms)
+        seen_signatures: set[tuple] = set()
+        unique_events = []
+        for event in all_events:
+            sig = (
+                event.get("col"),
+                _stable_repr(event.get("old_val")),
+                _stable_repr(event.get("new_val")),
+                event.get("operation"),
+            )
+            if sig not in seen_signatures:
+                seen_signatures.add(sig)
+                unique_events.append(event)
+
+        return unique_events
 
     def get_cell_history_with_lineage(
         self, row_id: int, column: str, max_depth: int = 10

{tracepipe-0.3.2 → tracepipe-0.3.4}/uv.lock

@@ -2051,7 +2051,7 @@ wheels = [
 
 [[package]]
 name = "tracepipe"
-version = "0.3.0"
+version = "0.3.2"
 source = { editable = "." }
 dependencies = [
     { name = "numpy", version = "2.0.2", source = { registry = "https://pypi.org/simple" }, marker = "python_full_version < '3.10'" },