You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I tried to train an xgboost model, I used a csv file without header line that is output of an all number pandas data-frame but got error: Alphabet found in the header line.
Then after a while I realized that it was because there is scientific notation in the csv file, e.g., 1e-10
It would be better if it can automatically convert scientific notation to a number. Also this problem is hard to debug, because I am sure that my csv file only contains numbers but the error basically says you have non-numbers.
Best,
Hantian
The text was updated successfully, but these errors were encountered:
HI @hantianz thanks for using SageMaker and pointing this out.
We look at all enhancement / feature requests and consider them as part of our backlog and are constantly looking at our prioirities. I agree providing a better error message or even just converting it would be great!
We will keep this issue open to track it.
apacker
pushed a commit
to apacker/sagemaker-python-sdk
that referenced
this issue
Nov 15, 2018
When I tried to train an xgboost model, I used a csv file without header line that is output of an all number pandas data-frame but got error: Alphabet found in the header line.
Then after a while I realized that it was because there is scientific notation in the csv file, e.g., 1e-10
It would be better if it can automatically convert scientific notation to a number. Also this problem is hard to debug, because I am sure that my csv file only contains numbers but the error basically says you have non-numbers.
Best,
Hantian
The text was updated successfully, but these errors were encountered: