Keep dashboard category identity separate from shortened labels

great_ai/deploy/routes/dashboard/create_dash_app.py

@@ -272,10 +272,10 @@ def get_dimension_descriptor(df: pd.DataFrame, column: str) -> Dict[str, Any]:
     except (TypeError, ValueError):
         MAX_LENGTH = 40
         unique_values = unique(values, key=freeze)
-        value_mapping = {str(v)[:MAX_LENGTH]: i for i, v in enumerate(unique_values)}
+        value_mapping = {freeze(v): i for i, v in enumerate(unique_values)}
 
-        dimension["values"] = [value_mapping[str(v)[:MAX_LENGTH]] for v in values]
+        dimension["values"] = [value_mapping[freeze(v)] for v in values]
         dimension["tickvals"] = list(value_mapping.values())
-        dimension["ticktext"] = [k[:MAX_LENGTH] for k in value_mapping.keys()]
+        dimension["ticktext"] = [str(value)[:MAX_LENGTH] for value in unique_values]
 
     return dimension

tests/test_dashboard_regressions.py

@@ -1,7 +1,8 @@
 from typing import Any
 
+import pandas as pd
 import pytest
-
+from great_ai.deploy.routes.dashboard.create_dash_app import get_dimension_descriptor
 from great_ai.deploy.routes.dashboard.get_filter_from_datatable import (
     get_filter_from_datatable,
 )
@@ -30,3 +31,13 @@ def test_dashboard_filter_grammar(expression: str, operator: str, value: Any) ->
 )
 def test_incomplete_filters_are_ignored(expression: str) -> None:
     assert get_filter_from_datatable(expression) is None
+
+
+def test_long_category_labels_keep_distinct_coordinates() -> None:
+    prefix = "a" * 40
+    dimension = get_dimension_descriptor(
+        pd.DataFrame({"category": [prefix + "1", prefix + "2"]}), "category"
+    )
+    assert dimension["values"] == [0, 1]
+    assert dimension["tickvals"] == [0, 1]
+    assert dimension["ticktext"] == [prefix, prefix]