[
https://issues.apache.org/jira/browse/SPARK-37449?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Carlos Gameiro updated SPARK-37449:
-----------------------------------
Description:
I'm using pygeos 0.11.1.
Let's create a simple Pandas Dataframe with a single column named 'id' with a
range.
{code:java}
df = pd.DataFrame(np.arange(0,1000), columns=['id']){code}
Consider this simple function that selects the first 4 indexes of the 'id'
column of an array.
{code:java}
def udf_example(df):
some_index = np.array([0, 1, 2, 3])
values = df['id'].values[some_index]
df = pd.DataFrame(values, columns=['id'])
return df{code}
If I apply this function in Pyspark I get this result:
{code:java}
schema = t.StructType([t.StructField('id', t.LongType(), True)])
df_spark = spark.createDataFrame(df).groupBy().applyInPandas(udf_example,
schema)
display(df_spark)
# id
# 125
# 126
# 127
# 128
{code}
If I apply it in Python I get the correct and expected result:
{code:java}
udf_example(df)
# id
# 0
# 1
# 2
# 3
{code}
Using NumPy indexing operations inside a Pandas UDF in Spark causes side
effects and unexpected results.
was:
I'm using pygeos 0.11.1.
Let's create a simple Pandas Dataframe with a single column named 'id' with a
range.
{code:java}
df = pd.DataFrame(np.arange(0,1000), columns=['id']){code}
Consider this simple function that selects the first 4 indexes of the 'id'
column of an array.
{code:java}
def udf_example(df):
some_index = np.array([0, 1, 2, 3])
values = df['id'].values[some_index]
df = pd.DataFrame(values, columns=['id'])
return df{code}
If I apply this function in Pyspark I get this result:
{code:java}
schema = t.StructType([t.StructField('id', t.LongType(), True)])
df_spark = spark.createDataFrame(df).groupBy().applyInPandas(udf_example,
schema)
display(df_spark)
# id
# 125
# 126
# 127
# 128
{code}
If I apply it in Python I get the correct and expected result:
{code:java}
udf_example(df)
# id
# 0
# 1
# 2
# 3
{code}
Using NumPy indexing operations inside a UDF in Spark causes side effects and
unexpected results.
> Side effects between PySpark and Numpy
> --------------------------------------
>
> Key: SPARK-37449
> URL: https://issues.apache.org/jira/browse/SPARK-37449
> Project: Spark
> Issue Type: Bug
> Components: PySpark
> Affects Versions: 3.1.2
> Reporter: Carlos Gameiro
> Priority: Critical
> Labels: NumPy, Pandas, Pygeos, UDF, applyInPandas
>
> I'm using pygeos 0.11.1.
> Let's create a simple Pandas Dataframe with a single column named 'id' with a
> range.
> {code:java}
> df = pd.DataFrame(np.arange(0,1000), columns=['id']){code}
> Consider this simple function that selects the first 4 indexes of the 'id'
> column of an array.
> {code:java}
> def udf_example(df):
>
> some_index = np.array([0, 1, 2, 3])
> values = df['id'].values[some_index]
>
> df = pd.DataFrame(values, columns=['id'])
> return df{code}
> If I apply this function in Pyspark I get this result:
> {code:java}
> schema = t.StructType([t.StructField('id', t.LongType(), True)])
> df_spark = spark.createDataFrame(df).groupBy().applyInPandas(udf_example,
> schema)
> display(df_spark)
> # id
> # 125
> # 126
> # 127
> # 128
> {code}
> If I apply it in Python I get the correct and expected result:
> {code:java}
> udf_example(df)
> # id
> # 0
> # 1
> # 2
> # 3
> {code}
> Using NumPy indexing operations inside a Pandas UDF in Spark causes side
> effects and unexpected results.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
