README.md

In ClaraMarquardt/featuristic: Generate ehR-based feature sets.

featuristic

NOTE: THIS VERSION OF THE REPOSITORY IS NO LONGER ACTIVELY MAINTAINED

Please see https://github.com/sysmedlab/featuristic

Install

library("devtools")  
install_github("claramarquardt/featuristic",dependencies = TRUE)    
library(featuristic)

Documentation - See function_overview.csv for an overview of all the functions (and data sets) included in the package - Usaging the package

1. Generate a local copy of the control.R template located at featuristic/template/control.R. 
2. Modify the settings contained in the control.R file (comments provided)
3. *Construct features - Stage I*
    feature_construction([path to local version of the control.R file])
3. *Compile features - Stage II*
    feature_compilation([path to local version of the control.R file])

Datasets - gagne_code: Icd9 code - gagne comorbidity category crosswalk (http://scholar.harvard.edu/gagne/software/combined-comorbidity-score) [ note the crosswalk included in the package includes gagne categories that are assigned a 0 weight (these are not included in the here referenced, publicly available version of the crosswalk)] - zip_class: US zip code - city/state crosswalk - flu_data_cdc: CDC city/state level flu data - variable_list_default*: Default feature groupings, i.e. feature subsets

Note - Requires the newest version of devtools (https://github.com/hadley/devtools) to be installed

library(devtools)  
install_github("hadley/devtools")

Development - Package is actively being developed and extended

• To contribute:

# 1. Clone the repo
git clone https://github.com/ClaraMarquardt/featuristic.git

# 2. Create a new branch
git checkout -b [branch name]

# 3. Push all changes to the branch (assuming all changes have been committed)
git push origin [branch name]

# 4. Test by installing from the branch
library(devtools)
install_git("git://github.com/ClaraMarquardt/featuristic.git", branch = "[branch name])

ClaraMarquardt/featuristic documentation built on May 9, 2019, 5:59 p.m.