Skip to content

Conversation

@VesnaT
Copy link
Contributor

@VesnaT VesnaT commented May 7, 2021

Issue

Fixes #5424

Description of changes

Add a check box to normalize/not normalize the input dataset. The check box is checked by default.

image

Includes
  • Code changes
  • Tests
  • Documentation

@codecov
Copy link

codecov bot commented May 7, 2021

Codecov Report

Merging #5428 (5a0f03e) into master (ce6222e) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #5428      +/-   ##
==========================================
- Coverage   86.14%   86.14%   -0.01%     
==========================================
  Files         303      303              
  Lines       61592    61607      +15     
==========================================
+ Hits        53061    53072      +11     
- Misses       8531     8535       +4     

@markotoplak
Copy link
Member

@VesnaT, thanks, looks good to me!

@markotoplak markotoplak merged commit 4376a88 into biolab:master May 7, 2021
@borondics
Copy link
Member

borondics commented May 7, 2021

This is a nice addition, but what about the other preprocessors in DBSCAN? The user still doesn't know about them, even if they are necessary...

The GUI across widgets is not consistent. I like k-means as there is a Preprocessing group, so one could add everything there and in DBSCAN's case the Distance metric could continue having its own group.
image

@markotoplak
Copy link
Member

The other preprocessors are necessary if you do not want the widget to work at all with missing values or some discrete columns. If your data does not have either they will not do anything anyway.

The UI's are different because one allows distance measurement method setting and the other does not.

@janezd
Copy link
Contributor

janezd commented May 7, 2021

But preprocessors should be listed in documentation, I guess. For this widget and for others.

@irgolic
Copy link
Member

irgolic commented May 7, 2021

Could we automate/standardize preprocessors, to make them a bit more transparent?

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.

Unknown Preprocessing Steps

5 participants