fix: remove the duplicated cardinality threshold under categorical an…

…d text settings
ydataai · Oct 10, 2023 · 80a1024 · 80a1024
1 parent f4886a2
commit 80a1024
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 6 deletions.
diff --git a/src/ydata_profiling/config.py b/src/ydata_profiling/config.py
@@ -50,17 +50,16 @@ class TextVars(BaseModel):
     words: bool = True
     characters: bool = True
     redact: bool = False
-    # if text has more than threshold categories, its not category
-    categorical_threshold: int = 50
-    # if text has more than threshold % distinct values, its not category
-    percentage_cat_threshold: float = 0.5
 
 
 class CatVars(BaseModel):
     length: bool = True
     characters: bool = True
     words: bool = True
+    # if var has more than threshold categories, it's a text var
     cardinality_threshold: int = 50
+    # if var has more than threshold % distinct values, it's a text var
+    percentage_cat_threshold: float = 0.5
     imbalance_threshold: float = 0.5
     n_obs: int = 5
     # Set to zero to disable

diff --git a/src/ydata_profiling/model/typeset_relations.py b/src/ydata_profiling/model/typeset_relations.py
@@ -73,8 +73,8 @@ def string_is_category(series: pd.Series, state: dict, k: Settings) -> bool:
     - (distinct values / count of all values) is less than threshold
     - is not bool"""
     n_unique = series.nunique()
-    unique_threshold = k.vars.text.percentage_cat_threshold
-    threshold = k.vars.text.categorical_threshold
+    unique_threshold = k.vars.cat.percentage_cat_threshold
+    threshold = k.vars.cat.cardinality_threshold
     return (
         1 <= n_unique <= threshold
         and (