Lukas Heumos

### Description of feature https://github.com/DynamicTimeWarping/dtw-python

Add time based sankey diagrams

3

Add a new type of plot. Sankey diagrams based on transitions such as time. We might have to use Plotly for this: https://plotly.com/python/sankey-diagram/

### Description of feature Calculate it automatically. Would allow us to reduce the number of methods

Support branching for cohort tracking

1

### Description of feature https://github.com/adayim/consort supports branching off. We should also enable that somehow.

Check out MICE extension for MNAR data

1

https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/s12874-020-01180-y

Integrate Synthea or create own fake data generator

6

https://github.com/synthetichealth/synthea We should try to wrap it in Python (while checking for JDK etc). Output default csv, but others should be supported as well. CC #102

Add module for causal discovery

1

We recently implemented causal inference with dowhy. However, currently ehrapy has no module to also conduct causal discovery or in other words determine causal graphs on its own. It'd be...

Two-sample Kolmogorov-Smirnov test to compare pre and post imputation

8

### Description of feature We could implement the two-sample Kolmogorov-Smirnov test which we could use to test pre and post imputation distributions > We tested the null hypothesis considering that...

Master issue to make everything work with sparse matrices

1

### Description of feature - [ ] Quality control - [ ] normalization - [ ] encoding -> #650 - [ ] [...]

Allow `ep.pp.encode` to produce sparse output

1

### Description of feature I looked a bit into sparse encoding. One-hot encoding being the most important: 1. scikit-learn's one-hot encoding supports a `sparse_output` parameter that should return a CSR...