Created
July 23, 2021 01:42
-
-
Save kierisi/b6be49677cf6243135bdc2c1aebb9a5b to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# setup ------------------------------------------------------------------- | |
library(tidymodels) | |
library(tidyverse) | |
library(baguette) | |
library(janitor) | |
library(DataExplorer) | |
library(here) | |
tidymodels_prefer() | |
options(tidymodels.dark = TRUE) | |
grid_control <- control_grid(save_pred = TRUE, | |
save_workflow = TRUE, | |
verbose = TRUE, | |
extract = extract_model) | |
# import and inspect ------------------------------------------------------ | |
path <- "05_lap_02_episode_08" | |
train <- read_csv(here(path, "train.csv")) | |
test <- read_csv(here(path, "test.csv")) | |
glimpse(train) | |
plot_intro(train) | |
plot_histogram(train) | |
plot_bar(train) | |
# wrangling --------------------------------------------------------------- | |
#' AT MINIMUM RUN THIS: | |
train_tidy <- train %>% | |
clean_names() %>% | |
mutate(release_month = as.numeric(release_month), | |
release_day = as.numeric(release_day)) %>% | |
select(id, popularity, duration_ms, danceability:release_day) | |
glimpse(train_tidy) | |
test_tidy <- test %>% | |
clean_names() %>% | |
mutate(release_month = as.numeric(release_month), | |
release_day = as.numeric(release_day)) %>% | |
select(id, duration_ms, danceability:release_day) | |
train_min <- train %>% | |
clean_names() %>% | |
select(id, popularity, release_year) | |
test_min <- test %>% | |
clean_names() %>% | |
select(id, release_year) | |
# RNG model --------------------------------------------------------------- | |
set.seed(406) | |
lol <- as_tibble(runif(9000)) | |
glimpse(lol) | |
glimpse(test) | |
wtf <- test %>% | |
select(id) %>% | |
bind_cols(lol) %>% | |
rename(popularity = value) | |
glimpse(wtf) | |
wtf %>% | |
write_csv(here(path, "lolwhy.csv")) | |
# spend our data ---------------------------------------------------------- | |
set.seed(406) | |
folds <- vfold_cv(train_tidy, v = 5) | |
folds_grid <- vfold_cv(train_tidy, v = 10, repeats = 5) | |
# bagged tree model ------------------------------------------------------- | |
## WTF got the weirdest error?!?! | |
bb_bt <- recipe(popularity ~ ., | |
data = train_tidy) %>% | |
step_novel(all_nominal_predictors()) %>% | |
update_role(id, new_role = "id") | |
bb_bt_spec <- bag_tree(min_n = 10) %>% | |
set_engine('rpart', times = 25) %>% | |
set_mode('regression') | |
bb_bt_workflow <- workflow() %>% | |
add_recipe(bb_bt) %>% | |
add_model(bb_bt_spec) | |
bb_bt_res <- fit_resamples( | |
bb_bt_workflow, | |
folds, | |
control = grid_control | |
) | |
collect_metrics(bb_bt_res) | |
bb_bt_fit <- fit(bb_bt_workflow, train_tidy) | |
bb_bt_pred <- augment(bb_bt_fit, new_data = test_tidy) | |
glimpse(bb_bt_pred) | |
bb_bt_pred %>% | |
# CHECK FIRST | |
#select(id, | |
#attrition_flag = .pred_1) %>% | |
write_csv(here(path, "bb_bt_01.csv")) | |
# troubleshooting -------------------------------------------------------- | |
#' error message: | |
#' 1 preprocessor 1/1, model 1/1: Error: indicators should be a character, | |
#' not a logical. | |
bb_bt %>% | |
prep() %>% | |
glimpse() | |
juice() %>% | |
glimpse() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment