More than 3 years have passed since last update.

R: Kaggle House Prices

Last updated at 2020-09-06Posted at 2020-09-06

こちらと同じことを、R で行いました。しかし、スコアはかなり違います。
scikit-learn の使い方 (その4)

[1]

library('randomForest')

[2]

train_path <- '../input/house-prices-advanced-regression-techniques/train.csv'
test_path <- '../input/house-prices-advanced-regression-techniques/test.csv'
submission_path <- '../input/house-prices-advanced-regression-techniques/sample_submission.csv'

train <- read.csv(train_path, stringsAsFactors = F )
test  <- read.csv(test_path, stringsAsFactors = F )
submission  <- read.csv(submission_path, stringsAsFactors = F )
#
train <- train[,c(5,18,20,55,81)]
test <- test[,c(5,18,20,55)]
#
names(train)
head(train)

head(submission)

[3]

set.seed(754)
rf_model <- randomForest(SalePrice ~ .,　data =  train)
print(rf_model)

[4]

prediction <- predict(rf_model, test)

[5]

submission.SalePrice = prediction
#
write.csv(submission, file = 'house_price_r_sep0601.csv', row.names = F)
head(prediction)
head(submission)
nrow(submission)
length(submission)
dim(submission)