0

I am trying to read a feather file into Python using pyarrow like so:

import pyarrow.feather as feather
feather.read_feather("../data/my_file.feather")

However, when I run this, I get the following error message:

ArrowInvalid: Unrecognized compression type: LZ4

According to the documentation here, LZ4 is supported for V2. I also tried using the feather package to read it into R, but there it tells me that the file is not a feather file. Any idea why this is happening? Does this indicate that the file itself is corrupt?

I am using Python 3.8 and the latest version of pyarrow (2.0.0).

Thanks!

Annie Chen
  • 61
  • 7
  • LZ4 compression should indeed be supported. How did you create the Feather file? Would you be able to share a reproducible example? – joris Jan 04 '21 at 09:48
  • Thanks for the response, @joris. I did not create the file, but you can download the file [here](https://zenodo.org/record/4327607#.X_SskulKi3I) (`yt_metadata_helper.feather`). – Annie Chen Jan 05 '21 at 18:17
  • 2
    Thanks for that link! I investigated a bit, and this is a bug in pyarrow. Reading with pyarrow 1.0.0 should still work, and hopefully we can fix it in time for the upcoming pyarrow 3.0.0 release. I reported it as https://issues.apache.org/jira/browse/ARROW-11163 – joris Jan 07 '21 at 15:04

0 Answers0