A library containing useful EHR related functions for use within Python data analysis, especially within Jupyter notebooks.
When utilizing Jupyter notebooks for processing data and training models I found myself copying the same code between notebooks. This code consisted of steps to split my data, create a model, compute some metrics, etc.; and thus the notebooks became very long with little focus on the actual analysis. Therefore, this set of functions were created to allow for a focus on analysis and to abstract away the process of cleaning data and running models.
pip install ehr-functions
The documentation can be found at: https://fdabek1.github.io/ehr-functions/
from ehr_functions.features import demographics
import pandas as pd
df = pd.DataFrame({
'PatientID': [1, 2, 3, 4],
'PatientAge': [21, 35, 27, 24],
'PatientGender': ['M', 'F', 'M', 'F'],
'PatientCategory': ['A', 'B', 'C', 'A'],
})
dems = demographics.get_features(df)
print(dems.head())| PatientID | PatientAge | PatientGender | PatientCategory_A | PatientCategory_B | PatientCategory_C |
|---|---|---|---|---|---|
| 1 | 21 | 1 | 1 | 0 | 0 |
| 2 | 35 | 0 | 0 | 1 | 0 |
| 3 | 27 | 1 | 0 | 0 | 1 |
| 4 | 24 | 0 | 1 | 0 | 0 |