zhengruifeng opened a new pull request, #46281: URL: https://github.com/apache/spark/pull/46281
### What changes were proposed in this pull request? Cache `DataFrame.isStreaming` ### Why are the changes needed? In PS, `DataFrame.isStreaming` is used in the construction of `InternalFrame` https://github.com/apache/spark/blob/e01ac581f46aa595e66daf33fe92b56d1328bc78/python/pyspark/pandas/internal.py#L624 it might cause performance issues since a lot of `InternalFrame` will be built in even simple logic, such as ``` import cProfile, pstats import pyspark.pandas as ps df1 = ps.DataFrame({'lkey': ['foo', 'bar', 'baz', 'foo'], 'value': [1, 2, 3, 5]}, columns=['lkey', 'value']) cProfile.run("df1['value2'] = df1['value'] + 123 + 456", "/tmp/profile_results") pstats.Stats("/tmp/profile_results").sort_stats("cumtime").print_stats(.1) ``` before: ``` ncalls tottime percall cumtime percall filename:lineno(function) 1 0.000 0.000 0.120 0.120 {built-in method builtins.exec} 1 0.000 0.000 0.120 0.120 <string>:1(<module>) 27 0.000 0.000 0.113 0.004 /Users/ruifeng.zheng/Dev/spark/python/pyspark/sql/connect/client/core.py:1160(_analyze) 27 0.000 0.000 0.112 0.004 /Users/ruifeng.zheng/.dev/miniconda3/envs/spark_dev_311/lib/python3.11/site-packages/grpc/_channel.py:1161(__call__) 27 0.000 0.000 0.112 0.004 /Users/ruifeng.zheng/.dev/miniconda3/envs/spark_dev_311/lib/python3.11/site-packages/grpc/_channel.py:1124(_blocking) 27 0.110 0.004 0.110 0.004 {method 'next_event' of 'grpc._cython.cygrpc.SegregatedCall' objects} 2 0.000 0.000 0.108 0.054 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/base.py:318(__add__) 2 0.000 0.000 0.078 0.039 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/data_type_ops/num_ops.py:79(add) 2 0.000 0.000 0.074 0.037 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/base.py:210(wrapper) 24 0.000 0.000 0.064 0.003 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/internal.py:1435(copy) 24 0.000 0.000 0.063 0.003 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/internal.py:535(__init__) ``` after: ``` ncalls tottime percall cumtime percall filename:lineno(function) 1 0.000 0.000 0.082 0.082 {built-in method builtins.exec} 1 0.000 0.000 0.082 0.082 <string>:1(<module>) 2 0.000 0.000 0.081 0.041 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/base.py:318(__add__) 4 0.000 0.000 0.078 0.019 /Users/ruifeng.zheng/Dev/spark/python/pyspark/sql/connect/client/core.py:1160(_analyze) 4 0.000 0.000 0.077 0.019 /Users/ruifeng.zheng/.dev/miniconda3/envs/spark_dev_311/lib/python3.11/site-packages/grpc/_channel.py:1161(__call__) 4 0.000 0.000 0.077 0.019 /Users/ruifeng.zheng/.dev/miniconda3/envs/spark_dev_311/lib/python3.11/site-packages/grpc/_channel.py:1124(_blocking) 4 0.077 0.019 0.077 0.019 {method 'next_event' of 'grpc._cython.cygrpc.SegregatedCall' objects} 26 0.000 0.000 0.058 0.002 /Users/ruifeng.zheng/Dev/spark/python/pyspark/sql/connect/dataframe.py:1783(schema) 3 0.000 0.000 0.058 0.019 /Users/ruifeng.zheng/Dev/spark/python/pyspark/sql/connect/client/core.py:1031(schema) 2 0.000 0.000 0.057 0.029 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/data_type_ops/num_ops.py:79(add) 2 0.000 0.000 0.057 0.028 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/base.py:210(wrapper) 24 0.000 0.000 0.026 0.001 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/internal.py:1435(copy) 21 0.000 0.000 0.025 0.001 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/internal.py:1419(select_column) 19 0.000 0.000 0.025 0.001 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/series.py:438(_internal) 24 0.000 0.000 0.025 0.001 /Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/internal.py:535(__init__) ``` There are 24 `InternalFrame.__init__` invocations in this example, the number of RPC is reduced from 27 to 4. ### Does this PR introduce _any_ user-facing change? no ### How was this patch tested? ci ### Was this patch authored or co-authored using generative AI tooling? no -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org
