add duplicate value warning, sthe order of those values will be arbitray

gseapy/base.py

@@ -135,6 +135,16 @@ def _load_ranking(self, rnk: Union[pd.DataFrame, pd.Series, str]) -> pd.Series:
             self._logger.warning("Input gene rankings contains inf values!")
             rankser.replace(-np.inf, method="ffill", inplace=True)
             rankser.replace(np.inf, method="bfill", inplace=True)
+
+        # check duplicate values and warning
+        dups = rankser.duplicated().sum()
+        if dups > 0:
+            msg = "Duplicated values found in preranked stats: {:.2%} of genes".format(
+                dups / rankser.size
+            )
+            msg += "The order of those genes will be arbitrary, which may produce unexpected results."
+            self._logger.warning(msg)
+
         # return series
         return rankser
 

gseapy/gsea.py

@@ -387,6 +387,16 @@ def load_ranking(self):
                 }  # column-wise min
                 ranking = ranking.replace({col: col_min_max for col in ranking.columns})
 
+            # check ties in prerank stats
+            dups = ranking.apply(lambda df: df.duplicated().sum() / df.size)
+            if (dups > 0).sum() > 0:
+                msg = (
+                    "Duplicated values found in preranked stats:\nsample\tratio\n%s\n"
+                    % (dups.to_string(float_format="{:,.2%}".format))
+                )
+                msg += "The order of those genes will be arbitrary, which may produce unexpected results."
+                self._logger.warning(msg)
+
         return ranking
 
     # @profile