| run_predict {alookr} | R Documentation |
Predict binary classification model
Description
Predict some representative binary classification models.
Usage
run_predict(model, .data, cutoff = 0.5)
Arguments
model |
A model_df. results of fitted model that created by run_models(). |
.data |
A tbl_df. The data set to predict the model. It also supports tbl, and data.frame objects. |
cutoff |
numeric. Cut-off that determines the positive from the probability of predicting the positive. |
Details
Supported models are functions supported by the representative model package used in R environment. The following binary classifications are supported:
"logistic" : logistic regression by predict.glm() in stats package.
"rpart" : recursive partitioning tree model by predict.rpart() in rpart package.
"ctree" : conditional inference tree model by predict() in stats package.
"randomForest" : random forest model by predict.randomForest() in randomForest package.
"ranger" : random forest model by predict.ranger() in ranger package.
"xgboost" : random forest model by predict.xgb.Booster() in xgboost package.
"lasso" : random forest model by predict.glmnet() in glmnet package.
run_predict() is executed in parallel when predicting by model. However, it is not supported in MS-Windows operating system and RStudio environment.
Value
model_df. results of predicted model. model_df is composed of tbl_df and contains the following variables.:
step : character. The current stage in the model fit process. The result of calling run_predict() is returned as "2.Predicted".
model_id : character. Type of fit model.
target : character. Name of target variable.
is_factor : logical. Indicates whether the target variable is a factor.
positive : character. Level of positive class of binary classification.
negative : character. Level of negative class of binary classification.
fitted_model : list. Fitted model object.
predicted : list. Predicted value by individual model. Each value has a predict_class class object.
Examples
library(dplyr)
# Divide the train data set and the test data set.
sb <- rpart::kyphosis %>%
split_by(Kyphosis)
# Extract the train data set from original data set.
train <- sb %>%
extract_set(set = "train")
# Extract the test data set from original data set.
test <- sb %>%
extract_set(set = "test")
# Sampling for unbalanced data set using SMOTE(synthetic minority over-sampling technique).
train <- sb %>%
sampling_target(seed = 1234L, method = "ubSMOTE")
# Cleaning the set.
train <- train %>%
cleanse
# Run the model fitting.
result <- run_models(.data = train, target = "Kyphosis", positive = "present")
result
# Run the several kinds model predict by dplyr
result %>%
run_predict(test)