scistackplotdb 0.1.48__tar.gz → 0.1.49__tar.gz

{scistackplotdb-0.1.48 → scistackplotdb-0.1.49}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: scistackplotdb
-Version: 0.1.48
+Version: 0.1.49
 Summary: scidb-backed plotting: load variables into long tables and plot them with scistackplot
 Author: SciStack Contributors
 License-Expression: MIT
@@ -17,10 +17,10 @@ Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Scientific/Engineering :: Visualization
 Classifier: Typing :: Typed
 Requires-Python: >=3.10
-Requires-Dist: sciduckdb==0.1.48
-Requires-Dist: scistack-db==0.1.48
-Requires-Dist: scistacklog==0.1.48
-Requires-Dist: scistackplot==0.1.48
+Requires-Dist: sciduckdb==0.1.49
+Requires-Dist: scistack-db==0.1.49
+Requires-Dist: scistacklog==0.1.49
+Requires-Dist: scistackplot==0.1.49
 Provides-Extra: dev
 Requires-Dist: matplotlib>=3.6; extra == 'dev'
 Requires-Dist: pytest-cov>=4.0; extra == 'dev'

{scistackplotdb-0.1.48 → scistackplotdb-0.1.49}/src/scistackplotdb/load.py

@@ -519,9 +519,6 @@ def load_variable(
         if columns and len(frame):
             with timer.phase("field_coverage"):
                 _log_field_coverage(variable, frame, column_types, levels)
-        if len(frame) > 1:
-            with timer.phase("identical_content"):
-                _log_identical_content(db, variable, frame, keys)
         return VariableFrame(
             name=variable,
             frame=frame,
@@ -684,56 +681,6 @@ def _log_field_coverage(
         Log.debug("field coverage of %s failed: %s", variable, exc, layer=LAYER)
 
 
-def _log_identical_content(db, variable: str, frame: pd.DataFrame, keys) -> None:
-    """Warn when records of ``variable`` hold byte-identical payloads, naming
-    the schema keys that vary across each duplicate group.
-
-    This is the one thing a plot cannot show you. N records with the same data
-    render as N traces drawn exactly on top of each other: the figure looks
-    like a single line in whatever colour was drawn last, and filtering to any
-    one value of the colour factor shows the same curve again. That is the
-    signature of a step that ran with ``distribute=False`` when it should have
-    distributed — every location of the key got the whole result instead of
-    its own slice — so it belongs in the log at load time rather than being
-    re-derived from a confusing figure.
-
-    Never raises: duplicate content is legal, and a diagnostic must not be able
-    to break a load.
-    """
-    try:
-        from scidb.provenance_query import identical_content_groups
-
-        groups = identical_content_groups(db._duck, variable)
-        if not groups:
-            return
-        varying: set = set()
-        indexed = frame.set_index("record_id")
-        for _hash, record_ids in groups:
-            present = [rid for rid in record_ids if rid in indexed.index]
-            if len(present) < 2:
-                continue
-            for key in keys:
-                values = {indexed.at[rid, key] for rid in present}
-                if len(values) > 1:
-                    varying.add(key)
-        Log.warn(
-            "%s: %d group(s) of records hold IDENTICAL data "
-            "(%d record(s) total); they differ only by %s. Plots colored or "
-            "faceted by those keys will draw overlapping identical traces. "
-            "This is what a step that should have run with distribute=True "
-            "looks like.",
-            variable,
-            len(groups),
-            sum(len(rids) for _h, rids in groups),
-            sorted(varying) or "no schema key",
-            layer=LAYER,
-        )
-    except Exception as e:  # diagnostics must never break a load
-        Log.debug(
-            "identical-content check skipped for %s: %s", variable, e, layer=LAYER
-        )
-
-
 def attach_variants(
     db, frame: pd.DataFrame
 ) -> tuple[pd.DataFrame, list[str], str | None, list[dict]]: