csvFromJSONs: Generate dataframes from a set of JSON Files

In IM-APHP/mlcodage: OpenCPU-based mlcodage package

Description

Generate 'df.text' and 'df.code' objects starting from a floder with a plenty of JSON files. These objects can then be given to function 'buildFromDF' in order to build a predictive model.

Usage

csvFromJSONs(jsFolder, diagsPath, corpusName, serviceName)

Arguments

jsFolder	path to your local folder containing JSON files.
diagsPath	path to your local RData file containing diagnostics (codes). The file must contain an object called "Diagnostics" with at least two columns ('NDA', and 'ACTE').
corpusName	choose a name for your corpus.
serviceName	choose a service name from the built-in services: {'urologie', 'chirgen', 'chirplas', 'chirmaxfac'}. This argument is used to choose the right UH (Hospital Unit) as well as the right variables from the JSON. If you don't want any restriction regarding UH and VAR, please set serviceName to 'gen'.

Value

a list of two dataframes: 'texts' and 'codes' to be given separately to function 'buildFromDF'.

IM-APHP/mlcodage documentation built on May 8, 2019, 10:52 a.m.