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

Feature - request inclusion of raw results as well as aggregated summaries in returned object #184

Open
py9mrg opened this issue Mar 22, 2022 · 1 comment

Comments

@py9mrg
Copy link

py9mrg commented Mar 22, 2022

Please consider including the raw results as well as aggregated results in the value returned by methods such as FeatureImp$new - e.g. in lines 312 - 325 of FeatureImp.R:

      result <- result[, list(
        "importance" = median(importance_raw),
        "permutation.error" = median(permutation_error),
        "importance.05" = quantile(importance_raw, probs = 0.05),
        "importance.95" = quantile(importance_raw, probs = 0.95)
      ), by = list(feature)]
      result <- result[order(result$importance, decreasing = TRUE), ]
      # Removes the n column
      result <- result[, list(
        feature, importance.05, importance, importance.95,
        permutation.error
      )]
      private$finished <- TRUE
      self$results <- data.frame(result)

It would be nice to add importance_raw as a list element either in results or at some higher level of the returned object. That way it allows us to make our own aggregations so we can change the cut off %s, plot the whole distribution, use a different summary statistic (e.g. I note another issue #170 is asking for mean instead of median). And similarly for other methods that return aggregated summaries from some raw initial result.

@py9mrg py9mrg changed the title Feature - request Feature - request inclusion of raw results as well as aggregated summaries in returned object Mar 22, 2022
@pat-s
Copy link
Collaborator

pat-s commented Mar 22, 2022

Thanks for contributing @py9mrg!

Note that this package is not actively developed right now. If you want to see such a change, then I suggest

  • you provide a complete reproducible example instead of a code excerpt
  • create a pull request which implements the proposed change in a first version

This will increase the chances getting it implemented, though there is no guarantee.

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

No branches or pull requests

2 participants