🐛 Fix ntile non-balanced bins by using pd.qcut, instead of `pd.cu…

…t` (pwwang/datar#161)
pwwang · Dec 12, 2022 · 7bd9a14 · 7bd9a14
1 parent 9907c30
commit 7bd9a14
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/datar_pandas/api/dplyr/_rank.py b/datar_pandas/api/dplyr/_rank.py
@@ -133,7 +133,7 @@ def _(x, n):
         return Categorical([np.nan] * x.size)
 
     n = min(n, x.size)
-    return pd.cut(x, n, labels=np.arange(n) + 1)
+    return pd.qcut(x, n, labels=np.arange(n) + 1)
 
 
 @_ntile.register(GroupBy)

diff --git a/datar_pandas/pandas.py b/datar_pandas/pandas.py
@@ -37,6 +37,7 @@
         crosstab,
         concat,
         cut,
+        qcut,
         isna,
         isnull,
         merge,
@@ -73,6 +74,7 @@ def get_obj(grouped):
         crosstab,
         concat,
         cut,
+        qcut,
         isna,
         isnull,
         merge,
-Original file line number
+Diff line change
@@ Expand Up / @@ -37,6 +37,7 @@ @@
             crosstab,
             concat,
             cut,
+            qcut,
             isna,
             isnull,
             merge,
@@ Expand Down Expand Up / @@ -73,6 +74,7 @@ def get_obj(grouped): @@
             crosstab,
             concat,
             cut,
+            qcut,
             isna,
             isnull,
             merge,
@@ Expand Down @@