My case of datasets stored is specific. I have large strings (1-100MB each).

Let's take for example a single row.

43mb.csv is a 1-row CSV with 10 columns. One column a 43mb string.

When I read this csv with pandas and then dump to parquet, my script consumes 10x of the 43mb.

With increasing amount of such rows memory footprint overhead diminishes, but I want to focus on this specific case.

Here's the footprint after running using memory profiler:

Line #    Mem usage    Increment   Line Contents
================================================
     4     48.9 MiB     48.9 MiB   @profile
     5                             def test():
     6    143.7 MiB     94.7 MiB       data = pd.read_csv('43mb.csv')
     7    498.6 MiB    354.9 MiB       data.to_parquet('out.parquet')
 

Is this typical for parquet in case of big strings?

Environment: Mac OSX
Reporter: Bogdan Klichuk

Related issues:

• [C++] Research jemalloc memory page reclamation configuration on macOS when background_thread option is unavailable (relates to)

Original Issue Attachments:

• 50mb.csv.gz

_{Note: This issue was originally created as ARROW-7305. Please see the migration documentation for further details.}