Upgrading PySpark#
Upgrading from PySpark 3.5 to 4.0#
In Spark 4.0, Python 3.8 support was dropped in PySpark.
In Spark 4.0, the minimum supported version for Pandas has been raised from 1.0.5 to 2.0.0 in PySpark.
In Spark 4.0, the minimum supported version for Numpy has been raised from 1.15 to 1.21 in PySpark.
In Spark 4.0, the minimum supported version for PyArrow has been raised from 4.0.0 to 10.0.0 in PySpark.
In Spark 4.0,
Int64Index
andFloat64Index
have been removed from pandas API on Spark,Index
should be used directly.In Spark 4.0,
DataFrame.iteritems
has been removed from pandas API on Spark, useDataFrame.items
instead.In Spark 4.0,
Series.iteritems
has been removed from pandas API on Spark, useSeries.items
instead.In Spark 4.0,
DataFrame.append
has been removed from pandas API on Spark, useps.concat
instead.In Spark 4.0,
Series.append
has been removed from pandas API on Spark, useps.concat
instead.In Spark 4.0,
DataFrame.mad
has been removed from pandas API on Spark.In Spark 4.0,
Series.mad
has been removed from pandas API on Spark.In Spark 4.0,
na_sentinel
parameter fromIndex.factorize
andSeries.factorize
has been removed from pandas API on Spark, useuse_na_sentinel
instead.In Spark 4.0,
inplace
parameter fromCategorical.add_categories
,Categorical.remove_categories
,Categorical.set_categories
,Categorical.rename_categories
,Categorical.reorder_categories
,Categorical.as_ordered
,Categorical.as_unordered
have been removed from pandas API on Spark.In Spark 4.0,
inplace
parameter fromCategoricalIndex.add_categories
,CategoricalIndex.remove_categories
,CategoricalIndex.remove_unused_categories
,CategoricalIndex.set_categories
,CategoricalIndex.rename_categories
,CategoricalIndex.reorder_categories
,CategoricalIndex.as_ordered
,CategoricalIndex.as_unordered
have been removed from pandas API on Spark.In Spark 4.0,
closed
parameter fromps.date_range
has been removed from pandas API on Spark.In Spark 4.0,
include_start
andinclude_end
parameters fromDataFrame.between_time
have been removed from pandas API on Spark, useinclusive
instead.In Spark 4.0,
include_start
andinclude_end
parameters fromSeries.between_time
have been removed from pandas API on Spark, useinclusive
instead.In Spark 4.0, the various datetime attributes of
DatetimeIndex
(day
,month
,year
etc.) are nowint32
instead ofint64
from pandas API on Spark.In Spark 4.0,
sort_columns
parameter fromDataFrame.plot
and Series.plot` has been removed from pandas API on Spark.In Spark 4.0, the default value of
regex
parameter forSeries.str.replace
has been changed fromTrue
toFalse
from pandas API on Spark. Additionally, a single characterpat
withregex=True
is now treated as a regular expression instead of a string literal.In Spark 4.0, the resulting name from
value_counts
for all objects sets to'count'
(or'proportion'
ifnormalize=True
was passed) from pandas API on Spark, and the index will be named after the original object.In Spark 4.0,
squeeze
parameter fromps.read_csv
andps.read_excel
has been removed from pandas API on Spark.In Spark 4.0,
null_counts
parameter fromDataFrame.info
has been removed from pandas API on Spark, useshow_counts
instead.In Spark 4.0, the result of
MultiIndex.append
does not keep the index names from pandas API on Spark.In Spark 4.0,
DataFrameGroupBy.agg
with lists respectingas_index=False
from pandas API on Spark.In Spark 4.0,
DataFrame.stack
guarantees the order of existing columns instead of sorting them lexicographically from pandas API on Spark.In Spark 4.0,
True
orFalse
toinclusive
parameter fromSeries.between
has been removed from pandas API on Spark, useboth
orneither
instead respectively.In Spark 4.0,
Index.asi8
has been removed from pandas API on Spark, useIndex.astype
instead.In Spark 4.0,
Index.is_type_compatible
has been removed from pandas API on Spark, useIndex.isin
instead.In Spark 4.0,
col_space
parameter fromDataFrame.to_latex
andSeries.to_latex
has been removed from pandas API on Spark.In Spark 4.0,
DataFrame.to_spark_io
has been removed from pandas API on Spark, useDataFrame.spark.to_spark_io
instead.In Spark 4.0,
Series.is_monotonic
andIndex.is_monotonic
have been removed from pandas API on Spark, useSeries.is_monotonic_increasing
orIndex.is_monotonic_increasing
instead respectively.In Spark 4.0,
DataFrame.get_dtype_counts
has been removed from pandas API on Spark, useDataFrame.dtypes.value_counts()
instead.In Spark 4.0,
encoding
parameter fromDataFrame.to_excel
andSeries.to_excel
have been removed from pandas API on Spark.In Spark 4.0,
verbose
parameter fromDataFrame.to_excel
andSeries.to_excel
have been removed from pandas API on Spark.In Spark 4.0,
mangle_dupe_cols
parameter fromread_csv
has been removed from pandas API on Spark.In Spark 4.0,
DataFrameGroupBy.backfill
has been removed from pandas API on Spark, useDataFrameGroupBy.bfill
instead.In Spark 4.0,
DataFrameGroupBy.pad
has been removed from pandas API on Spark, useDataFrameGroupBy.ffill
instead.In Spark 4.0,
Index.is_all_dates
has been removed from pandas API on Spark.In Spark 4.0,
convert_float
parameter fromread_excel
has been removed from pandas API on Spark.In Spark 4.0,
mangle_dupe_cols
parameter fromread_excel
has been removed from pandas API on Spark.In Spark 4.0,
DataFrame.koalas
has been removed from pandas API on Spark, useDataFrame.pandas_on_spark
instead.In Spark 4.0,
DataFrame.to_koalas
has been removed from PySpark, useDataFrame.pandas_api
instead.In Spark 4.0,
DataFrame.to_pandas_on_spark
has been removed from PySpark, useDataFrame.pandas_api
instead.In Spark 4.0,
DatatimeIndex.week
andDatatimeIndex.weekofyear
have been removed from Pandas API on Spark, useDatetimeIndex.isocalendar().week
instead.In Spark 4.0,
Series.dt.week
andSeries.dt.weekofyear
have been removed from Pandas API on Spark, useSeries.dt.isocalendar().week
instead.In Spark 4.0, when applying
astype
to a decimal type object, the existing missing value is changed toTrue
instead ofFalse
from Pandas API on Spark.In Spark 4.0,
pyspark.testing.assertPandasOnSparkEqual
has been removed from Pandas API on Spark, usepyspark.pandas.testing.assert_frame_equal
instead.In Spark 4.0, the aliases
Y
,M
,H
,T
,S
have been deprecated from Pandas API on Spark, useYE
,ME
,h
,min
,s
instead respectively.In Spark 4.0, the schema of a map column is inferred by merging the schemas of all pairs in the map. To restore the previous behavior where the schema is only inferred from the first non-null pair, you can set
spark.sql.pyspark.legacy.inferMapTypeFromFirstPair.enabled
totrue
.In Spark 4.0, compute.ops_on_diff_frames is on by default. To restore the previous behavior, set compute.ops_on_diff_frames to false.
Upgrading from PySpark 3.3 to 3.4#
In Spark 3.4, the schema of an array column is inferred by merging the schemas of all elements in the array. To restore the previous behavior where the schema is only inferred from the first element, you can set
spark.sql.pyspark.legacy.inferArrayTypeFromFirstElement.enabled
totrue
.In Spark 3.4, if Pandas on Spark API
Groupby.apply
’sfunc
parameter return type is not specified andcompute.shortcut_limit
is set to 0, the sampling rows will be set to 2 (ensure sampling rows always >= 2) to make sure infer schema is accurate.In Spark 3.4, if Pandas on Spark API
Index.insert
is out of bounds, will raise IndexError withindex {} is out of bounds for axis 0 with size {}
to follow pandas 1.4 behavior.In Spark 3.4, the series name will be preserved in Pandas on Spark API
Series.mode
to follow pandas 1.4 behavior.In Spark 3.4, the Pandas on Spark API
Index.__setitem__
will first to checkvalue
type isColumn
type to avoid raising unexpectedValueError
inis_list_like
like Cannot convert column into bool: please use ‘&’ for ‘and’, ‘|’ for ‘or’, ‘~’ for ‘not’ when building DataFrame boolean expressions..In Spark 3.4, the Pandas on Spark API
astype('category')
will also refreshcategories.dtype
according to original datadtype
to follow pandas 1.4 behavior.In Spark 3.4, the Pandas on Spark API supports groupby positional indexing in
GroupBy.head
andGroupBy.tail
to follow pandas 1.4. Negative arguments now work correctly and result in ranges relative to the end and start of each group, Previously, negative arguments returned empty frames.In Spark 3.4, the infer schema process of
groupby.apply
in Pandas on Spark, will first infer the pandas type to ensure the accuracy of the pandasdtype
as much as possible.In Spark 3.4, the
Series.concat
sort parameter will be respected to follow pandas 1.4 behaviors.In Spark 3.4, the
DataFrame.__setitem__
will make a copy and replace pre-existing arrays, which will NOT be over-written to follow pandas 1.4 behaviors.In Spark 3.4, the
SparkSession.sql
and the Pandas on Spark APIsql
have got new parameterargs
which provides binding of named parameters to their SQL literals.In Spark 3.4, Pandas API on Spark follows for the pandas 2.0, and some APIs were deprecated or removed in Spark 3.4 according to the changes made in pandas 2.0. Please refer to the [release notes of pandas](https://pandas.pydata.org/docs/dev/whatsnew/) for more details.
In Spark 3.4, the custom monkey-patch of
collections.namedtuple
was removed, andcloudpickle
was used by default. To restore the previous behavior for any relevant pickling issue ofcollections.namedtuple
, setPYSPARK_ENABLE_NAMEDTUPLE_PATCH
environment variable to1
.
Upgrading from PySpark 3.2 to 3.3#
In Spark 3.3, the
pyspark.pandas.sql
method follows [the standard Python string formatter](https://docs.python.org/3/library/string.html#format-string-syntax). To restore the previous behavior, setPYSPARK_PANDAS_SQL_LEGACY
environment variable to1
.In Spark 3.3, the
drop
method of pandas API on Spark DataFrame supports dropping rows byindex
, and sets dropping by index instead of column by default.In Spark 3.3, PySpark upgrades Pandas version, the new minimum required version changes from 0.23.2 to 1.0.5.
In Spark 3.3, the
repr
return values of SQL DataTypes have been changed to yield an object with the same value when passed toeval
.
Upgrading from PySpark 3.1 to 3.2#
In Spark 3.2, the PySpark methods from sql, ml, spark_on_pandas modules raise the
TypeError
instead ofValueError
when are applied to a param of inappropriate type.In Spark 3.2, the traceback from Python UDFs, pandas UDFs and pandas function APIs are simplified by default without the traceback from the internal Python workers. In Spark 3.1 or earlier, the traceback from Python workers was printed out. To restore the behavior before Spark 3.2, you can set
spark.sql.execution.pyspark.udf.simplifiedTraceback.enabled
tofalse
.In Spark 3.2, pinned thread mode is enabled by default to map each Python thread to the corresponding JVM thread. Previously, one JVM thread could be reused for multiple Python threads, which resulted in one JVM thread local being shared to multiple Python threads. Also, note that now
pyspark.InheritableThread
orpyspark.inheritable_thread_target
is recommended to use together for a Python thread to properly inherit the inheritable attributes such as local properties in a JVM thread, and to avoid a potential resource leak issue. To restore the behavior before Spark 3.2, you can setPYSPARK_PIN_THREAD
environment variable tofalse
.
Upgrading from PySpark 2.4 to 3.0#
In Spark 3.0, PySpark requires a pandas version of 0.23.2 or higher to use pandas related functionality, such as
toPandas
,createDataFrame
from pandas DataFrame, and so on.In Spark 3.0, PySpark requires a PyArrow version of 0.12.1 or higher to use PyArrow related functionality, such as
pandas_udf
,toPandas
andcreateDataFrame
with “spark.sql.execution.arrow.enabled=true”, etc.In PySpark, when creating a
SparkSession
withSparkSession.builder.getOrCreate()
, if there is an existingSparkContext
, the builder was trying to update theSparkConf
of the existingSparkContext
with configurations specified to the builder, but theSparkContext
is shared by allSparkSession
s, so we should not update them. In 3.0, the builder comes to not update the configurations. This is the same behavior as Java/Scala API in 2.3 and above. If you want to update them, you need to update them prior to creating aSparkSession
.In PySpark, when Arrow optimization is enabled, if Arrow version is higher than 0.11.0, Arrow can perform safe type conversion when converting pandas.Series to an Arrow array during serialization. Arrow raises errors when detecting unsafe type conversions like overflow. You enable it by setting
spark.sql.execution.pandas.convertToArrowArraySafely
to true. The default setting is false. PySpark behavior for Arrow versions is illustrated in the following table:PyArrow version
Integer overflow
Floating point truncation
0.11.0 and below
Raise error
Silently allows
> 0.11.0, arrowSafeTypeConversion=false
Silent overflow
Silently allows
> 0.11.0, arrowSafeTypeConversion=true
Raise error
Raise error
In Spark 3.0,
createDataFrame(..., verifySchema=True)
validates LongType as well in PySpark. Previously, LongType was not verified and resulted in None in case the value overflows. To restore this behavior, verifySchema can be set to False to disable the validation.As of Spark 3.0,
Row
field names are no longer sorted alphabetically when constructing with named arguments for Python versions 3.6 and above, and the order of fields will match that as entered. To enable sorted fields by default, as in Spark 2.4, set the environment variablePYSPARK_ROW_FIELD_SORTING_ENABLED
to true for both executors and driver - this environment variable must be consistent on all executors and driver; otherwise, it may cause failures or incorrect answers. For Python versions less than 3.6, the field names will be sorted alphabetically as the only option.In Spark 3.0,
pyspark.ml.param.shared.Has*
mixins do not provide anyset*(self, value)
setter methods anymore, use the respectiveself.set(self.*, value)
instead. See SPARK-29093 for details.
Upgrading from PySpark 2.3 to 2.4#
In PySpark, when Arrow optimization is enabled, previously
toPandas
just failed when Arrow optimization is unable to be used whereascreateDataFrame
from Pandas DataFrame allowed the fallback to non-optimization. Now, bothtoPandas
andcreateDataFrame
from Pandas DataFrame allow the fallback by default, which can be switched off byspark.sql.execution.arrow.fallback.enabled
.
Upgrading from PySpark 2.3.0 to 2.3.1 and above#
As of version 2.3.1 Arrow functionality, including
pandas_udf
andtoPandas()
/createDataFrame()
withspark.sql.execution.arrow.enabled
set toTrue
, has been marked as experimental. These are still evolving and not currently recommended for use in production.
Upgrading from PySpark 2.2 to 2.3#
In PySpark, now we need Pandas 0.19.2 or upper if you want to use Pandas related functionalities, such as
toPandas
,createDataFrame
from Pandas DataFrame, etc.In PySpark, the behavior of timestamp values for Pandas related functionalities was changed to respect session timezone. If you want to use the old behavior, you need to set a configuration
spark.sql.execution.pandas.respectSessionTimeZone
to False. See SPARK-22395 for details.In PySpark,
na.fill()
orfillna
also accepts boolean and replaces nulls with booleans. In prior Spark versions, PySpark just ignores it and returns the original Dataset/DataFrame.In PySpark,
df.replace
does not allow to omit value whento_replace
is not a dictionary. Previously, value could be omitted in the other cases and had None by default, which is counterintuitive and error-prone.
Upgrading from PySpark 1.4 to 1.5#
Resolution of strings to columns in Python now supports using dots (.) to qualify the column or access nested values. For example
df['table.column.nestedField']
. However, this means that if your column name contains any dots you must now escape them using backticks (e.g.,table.`column.with.dots`.nested
).DataFrame.withColumn method in PySpark supports adding a new column or replacing existing columns of the same name.
Upgrading from PySpark 1.0-1.2 to 1.3#
When using DataTypes in Python you will need to construct them (i.e.
StringType()
) instead of referencing a singleton.