Chris Kennedy

I believe I need to transform the continuous outcome estimates back to the original scale. Currently they're reported on the [0, 1] scale which is not ideal.

Error in knn imputation

1

There seems to be an error with knn imputation when a row has multiple missing values (possibly all missing values or all except 1 missing values). We may need an...

Via Stephanie Leonard, if there is a single variable in our X we get an error: `Error: invalid 'type' (list) of argument` We should 1) give a warning or error...

Dataset splitting fails due to "least populated class"

3

Hello, I've been analyzing a dataset and when pulling an updated version I now can no longer split the dataset using the darwin CLI. I receive this error: ``` Error:...

Fit treatment propensity based only on initial time point

8

Hello, I'm trying to conduct an intent-to-treat analysis and am having difficulty fitting the treatment propensity score using only t == 0 (baseline) and then using that estimated baseline treatment...

Add option for observation weights

2

Hello, I mentioned this in person but thought I'd start a quick feature request issue to support observation weights. I believe Mark is on board with simply passing the weights...

Parallelization support via future.apply

1

Hello, Here is a small patch to support parallel processing for test_consequences_data_frame(). When it takes longer to calculate those stats, such as larger dataframes, more granular thresholds, and/or more models,...

I think it would be good to upgrade the sentence segmentation to use spacy: https://spacy.io/usage/linguistic-features#sbd ```python nlp = spacy.load("en_core_web_sm") doc = nlp(u"This is a sentence. This is another sentence.") for...

Fix ranger.unify for probability forests

3

This is a minor fix to allow ranger.unify() to work correctly when probability = TRUE in a ranger::ranger() model, i.e. a probability forest. This example code will now work correctly:...