model_comparison_table: model_comparison_table Takes in training data, testing data,...

In UBC-MDS/RMLViz: Machine learning results visualization helper in R

Description

model_comparison_table Takes in training data, testing data, with the target as the last column and fitted models with meaningful names, then generates a model comparison table.

Usage

model_comparison_table(train_data, test_data, ...)

Arguments

train_data	tibble of training data with target as last column
test_data	tibble of testing data with target as last column
...	fitted models assigned to meaningful model names

Value

tibble of results, allowing you to compare models

Examples

## Not run: 
library(RMLViz)
library(mlbench)
data(Sonar)

toy_classification_data <- dplyr::select(dplyr::as_tibble(Sonar), V1, V2, V3, V4, V5, Class)

train_ind <- caret::createDataPartition(toy_classification_data$Class, p=0.9, list=F)
train_set_cf <- toy_classification_data[train_ind, ]
test_set_cf <- toy_classification_data[-train_ind, ]

gbm <- caret::train(Class~., train_set_cf, method="gbm", verbose=F)
lm_cf <- caret::train(Class~., train_set_cf, method="LogitBoost", verbose=F)

model_comparison_table(train_set_cf, test_set_cf,
                       gbm_mod=gbm, log_mod = lm_cf)

## End(Not run)

UBC-MDS/RMLViz documentation built on April 1, 2020, 4:35 a.m.