search

kowaalczyk / astronomical-classification

201st place (top 19%) submission for PLAsTiCC Astronomical Classification Kaggle Competition, team MIMUW E L I T E.
https://www.kaggle.com/c/PLAsTiCC-2018
4 stars 0 forks source link

Feature/training and cross validation #17

Closed radrow closed 5 years ago

radrow commented 6 years ago

fixes: #12

mkfilipiuk commented 6 years ago

Honestly, I don't really like the current implementation of dataset - it doesn't like wrapper for dataset, rather like pipeline to load it. I think we should store a pandas dataframe inside and add methods to extract X and Y without assuming in which column we store the Y Waiting for some sugestions 😃

radrow commented 6 years ago

@mkfilipiuk After some thinking I agree, it would be better to make dataset represent strict data instead of something to be loaded. I'll create an issue for this, but let's bring it as it is on the meeting

mkfilipiuk commented 6 years ago

Cool, see you in the evening :)

On Mon, Nov 5, 2018 at 10:43 AM Radosław Rowicki notifications@github.com wrote:

@mkfilipiuk https://github.com/mkfilipiuk After some thinking I agree, it would be better to make dataset represent strict data instead of something to be loaded. I'll create an issue for this, but let's bring it as it is on the meeting

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/kowaalczyk/astronomical-classification/pull/17#issuecomment-435813324, or mute the thread https://github.com/notifications/unsubscribe-auth/ARobVIEqPidyuME80FL7dxJo4yCgP2lCks5usAhZgaJpZM4YE35o .