Remove clear memory#150
Conversation
Codecov Report
@@ Coverage Diff @@
## master #150 +/- ##
==========================================
+ Coverage 57.77% 58.25% +0.47%
==========================================
Files 113 113
Lines 6960 6848 -112
==========================================
- Hits 4021 3989 -32
+ Misses 2939 2859 -80
|
kiudee
left a comment
There was a problem hiding this comment.
Looks good to me, thank you.
I agree with this simplification. Should it turn out that we need to clear memory in between different fit calls, we should anyway check what the best practices are.
|
Travis has succeeded but that information apparently somehow got lost in transit. Can we restart that check? |
|
Restarting the build somehow did not fix it. I think the only thing we could do is to push something to trigger the whole process again. If that does not work I will investigate if there is some connection issue between Github and Travis CI (so far there is no error message). |
Its use is no longer clear (kiudee#148) and it introduces unnecessary complications and additional state. Its removal also allows us to remove the hash_file parameter.
Reduces complexity and brings us one step closer to removing state (attributes) of estimators that is not set through the init parameters.
1e165d0 to
76a9854
Compare
|
Looks like this issue: |
|
Yeah I saw that too, very weird that it is just one report (all the previous ones seem outdated) and it isn't acknowledged yet. Anyway, thanks for the review! |
Description
Removing
clear_memoryandhash_fileto reduce complexity and get one step further towards stateless / scikit-learn compatible estimators.Motivation and Context
See description.
How Has This Been Tested?
Tests and pre-commit.
Does this close/impact existing issues?
Fixes #148, step towards #94 / #116.
Types of changes
Checklist: