[
https://issues.apache.org/jira/browse/ARROW-18400?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17642421#comment-17642421
]
Joris Van den Bossche commented on ARROW-18400:
-----------------------------------------------
While combining the chunks before converting to pandas is a useful workaround,
that still seems to point to a bug or inefficiency in the
pyarrow.Table->pandas.DataFrame conversion for nested columns.
The main conversion logic for this lives in {{arrow_to_pandas.cc}}
({{ConvertStruct}} and {{ConvertListLike}}). For structs, it seems to iterate
over the chunks of the ChunkedArray, and in that loop, iterate over the fields
to convert each field array to a numpy array and then create python
dictionaries combining those fields. And those dictionaries are inserted into
the numpy object dtype array for the full column (allocated in advance, for all
chunks).
So from that quick look, I don't directly see why combining the chunks in
advance would help / why having multiple chunks results in a higher memory
usage.
> [Python] Quadratic memory usage of Table.to_pandas with nested data
> -------------------------------------------------------------------
>
> Key: ARROW-18400
> URL: https://issues.apache.org/jira/browse/ARROW-18400
> Project: Apache Arrow
> Issue Type: Bug
> Components: Python
> Affects Versions: 10.0.1
> Environment: Python 3.10.8 on Fedora Linux 36. AMD Ryzen 9 5900 X
> with 64 GB RAM
> Reporter: Adam Reeve
> Assignee: Alenka Frim
> Priority: Critical
> Fix For: 11.0.0
>
>
> Reading nested Parquet data and then converting it to a Pandas DataFrame
> shows quadratic memory usage and will eventually run out of memory for
> reasonably small files. I had initially thought this was a regression since
> 7.0.0, but it looks like 7.0.0 has similar quadratic memory usage that kicks
> in at higher row counts.
> Example code to generate nested Parquet data:
> {code:python}
> import numpy as np
> import random
> import string
> import pandas as pd
> _characters = string.ascii_uppercase + string.digits + string.punctuation
> def make_random_string(N=10):
> return ''.join(random.choice(_characters) for _ in range(N))
> nrows = 1_024_000
> filename = 'nested.parquet'
> arr_len = 10
> nested_col = []
> for i in range(nrows):
> nested_col.append(np.array(
> [{
> 'a': None if i % 1000 == 0 else np.random.choice(10000,
> size=3).astype(np.int64),
> 'b': None if i % 100 == 0 else random.choice(range(100)),
> 'c': None if i % 10 == 0 else make_random_string(5)
> } for i in range(arr_len)]
> ))
> df = pd.DataFrame({'c1': nested_col})
> df.to_parquet(filename)
> {code}
> And then read into a DataFrame with:
> {code:python}
> import pyarrow.parquet as pq
> table = pq.read_table(filename)
> df = table.to_pandas()
> {code}
> Only reading to an Arrow table isn't a problem, it's the to_pandas method
> that exhibits the large memory usage. I haven't tested generating nested
> Arrow data in memory without writing Parquet from Pandas but I assume the
> problem probably isn't Parquet specific.
> Memory usage I see when reading different sized files on a machine with 64 GB
> RAM:
> ||Num rows||Memory used with 10.0.1 (MB)||Memory used with 7.0.0 (MB)||
> |32,000|362|361|
> |64,000|531|531|
> |128,000|1,152|1,101|
> |256,000|2,888|1,402|
> |512,000|10,301|3,508|
> |1,024,000|38,697|5,313|
> |2,048,000|OOM|20,061|
> |4,096,000| |OOM|
> With Arrow 10.0.1, memory usage approximately quadruples when row count
> doubles above 256k rows. With Arrow 7.0.0 memory usage is more linear but
> then quadruples from 1024k to 2048k rows.
> PyArrow 8.0.0 shows similar memory usage to 10.0.1 so it looks like something
> changed between 7.0.0 and 8.0.0.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
