From d251264c12f1a0dc2ac0aa61c609cddecd91244f Mon Sep 17 00:00:00 2001
From: Takuya UESHIN <ueshin@databricks.com>
Date: Fri, 13 Nov 2020 15:58:41 -0800
Subject: [PATCH] Fix to_csv to avoid duplicated option 'path' for
 DataFrameWriter. (#1912)

Spark doesn't allow duplicated option `path` when `DataFrameWriter.save()` since Spark 3.1. (apache/spark#29543)
---
 databricks/koalas/generic.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/databricks/koalas/generic.py b/databricks/koalas/generic.py
index 3b66b45823..c42bb8890b 100644
--- a/databricks/koalas/generic.py
+++ b/databricks/koalas/generic.py
@@ -828,7 +828,6 @@ def to_csv(
         if partition_cols is not None:
             builder.partitionBy(partition_cols)
         builder._set_opts(
-            path=path,
             sep=sep,
             nullValue=na_rep,
             header=header,