originalRidgeRegression: Linaer Regression Algorithm for Comparison
In WeakCha/BIOSTAT625_HW4: Linear Regression

Description Usage Arguments Value Author(s) Examples

View source: R/LinearRegression.R

Linaer Regression Algorithm for Comparison

1	originalRidgeRegression(data, features = NULL, target, training_part = 0.8, seed = 200)

`data`	`data` is a `data.frame`.
`training_part`	`training_part` is the ratio for training data.
`features`	`features` is a vector with the names of the features you want to use as predictors.
`target`	`target` is a string with the name of the target. We only allow uni-label predicting in this package.
`seed`	`seed` is the seed you want to specify, this ensures the reproducibility of the data.

`pred`	The prediction for the test set the algorithm automatically splitted
`RMSE`	The RMSE between the prediction and true target result for the test set

Li (Richard) Liu

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (data, features = NULL, target, training_part = 0.8,
    seed = 200)
{
    flag = require("glmnet")
    if (flag == FALSE) {
        install.packages("glmnet")
        library(glmnet)
    }
    set.seed(seed)
    index = sample(1:nrow(data), training_part * nrow(data))
    data_train = data[index, ]
    data_test = data[-index, ]
    X_train = data.matrix(data_train[, features])
    Y_train = data_train[, target]
    X_test = data.matrix(data_test[, features])
    Y_test = data_test[, target]
    model = glmnet(X_train, Y_train, alpha = 0, lambda = 0, family = "gaussian")
    pred = predict(model, X_test)
    pred = as.vector(pred)
    RMSE = eval_metrics(Y_test, as.vector(pred))
    pred = as.matrix(pred)
    return(list(pred = pred, RMSE = RMSE))
  }