Tools for Healthcare Machine Learning


[Up] [Top]

Documentation for package ‘healthcareai’ version 0.1.11

Help Pages

healthcareai-package healthcareai: a streamlined way to develop and deploy models
calculateAllCorrelations Correlation analysis on an input table over all numeric columns
calculatePerformance Generate performance metrics after model has been trained
calculateSDChanges Calculate std deviation up/down for each numeric field in row
calculateTargetedCorrelations Correlation analysis on an input table, focusing on one target variable
calulcateAlternatePredictions Recalculate predicted value based on alternate scenarios
convertDateTimeColToDummies Convert datetime column into dummy columns
countDaysSinceFirstDate Creates column based on days since first date
countPercentEmpty Calculates percentage of each column in df that is NULL (NA)
findBestAlternateScenarios Find most biggest drop in predictive probability across alternate features
findTrends Find any columns that have a trend above a particular threshold
generateAUC Generate ROC curve for a dataset.
groupedLOCF Last observation carried forward
healthcareai healthcareai: a streamlined way to develop and deploy models
imputeColumn Perform imputation on a vector
initializeParamsForTesting Function to initialize and populate the SupervisedModelDevelopmentParams each time a unit test is run.
isBinary Check if a vector has only two unique values.
LassoDeployment Deploy a production-ready predictive Lasso model
LassoDevelopment Compare predictive models, created on your data
LinearMixedModelDeployment Deploy a production-ready predictive Random Forest model
LinearMixedModelDevelopment Compare predictive models, created on your data
orderByDate Order the rows in a data frame by date
plotPRCurve Plot PR Curves from SupervisedModel classes
plotROCs Plot ROCs from SupervisedModel classes
RandomForestDeployment Deploy a production-ready predictive RandomForest model
RandomForestDevelopment Compare predictive models, created on your data
removeColsWithAllSameValue Remove columns from a data frame when those columns have the same values in each row
removeRowsWithNAInSpecCol Remove rows where specified col is NA
returnColsWithMoreThanFiftyCategories Return vector of columns in a data frame with greater than 50 categories
RiskAdjustedComparisons Make risk adjusted comparisons between groups/units or years/months
selectData Pull data into R via an ODBC connection
SupervisedModelDeployment Deploy predictive models, created on your data
SupervisedModelDeploymentParams SupervisedModelDeploymentParams class to set up parameters required to build SupervisedModelDeployment class
SupervisedModelDevelopment Compare predictive models, created on your data
SupervisedModelDevelopmentParams SupervisedModelDevelopmentParams class to set up parameters required to build SupervisedModel classes
writeData Write data to database