Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix: specify dtype for read_csv method #88

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

nielsbril
Copy link

@nielsbril nielsbril commented Sep 10, 2019

Fixing #87

@nielsbril
Copy link
Author

@JosseVanDelm @jbelien Can you have a look at this? We use my own fork for now, but it would be nice to use your official repo in the future.

@@ -62,7 +62,7 @@ def filter_file(args):
"""
logger.info('Started reading input file')
try:
file = pd.read_csv(args.input_file)
file = pd.read_csv(args.input_file, dtype='unicode')
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not a Python expert so I may be wrong but, since the script is supposed to be run with Python 3, shouldn't it be dtype='str' ?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@jbelien i just read this stackoverflow post:
https://stackoverflow.com/questions/24251219/pandas-read-csv-low-memory-and-dtype-options
According to this post the proposed code should silence the error, but it does not resolve the main problem.
pandas is trying to guess the datatype for every csv column, but to do this, it has to load in all the data in memory.

If i get it correctly, the proper way to do it is to explicitly state the numpy datatypes for each column in order to make the code more efficient. I am not sure what implications this has on the rest on the code, so for now we can definitely accept this push request, but we should further investigate on this issue in the future.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not a Python developer either, this code change seemed to fix the issue. It tells panda to treat each column as unicode, which resolves the issue for strings, numbers, ... The fix has been working for several months now on my own fork, but could (temporarily) be applied here too. But I agree this should be further investigated by someone with more knowledge on Python and the panda module.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants