Generic S3 Function for rpwf_augment() Object into the Database
Usage
rpwf_export_db(obj, db_con)
# S3 method for rpwf_workflow_set
rpwf_export_db(obj, db_con)
# S3 method for rpwf_data_set
rpwf_export_db(obj, db_con)Arguments
- obj
an augmented
rpwf_workflow_set()orrpwf_data_set()object.- db_con
an
rpwf_connect_db()object.
Examples
# Create the database
board <- pins::board_temp()
tmp_dir <- tempdir()
db_con <- rpwf_connect_db(paste(tmp_dir, "db.SQLite", sep = "/"), board)
# Create a `workflow_set`
d <- mtcars
d$target <- stats::rbinom(nrow(d), 1, 0.5)
m1 <- parsnip::boost_tree() |>
parsnip::set_engine("xgboost") |>
parsnip::set_mode("classification") |>
set_py_engine("xgboost", "XGBClassifier", "my_xgboost_model")
r1 <- d |>
recipes::recipe(target ~ .)
wf <- rpwf_workflow_set(list(r1), list(m1), "neg_log_loss")
to_export <- wf |>
rpwf_augment(db_con, dials::grid_latin_hypercube, size = 10)
#> No hyper param tuning specified
#> No pandas idx added. Use update_roles() with 'pd.index' for one
rpwf_write_grid(to_export)
#> No grid generated
rpwf_write_df(to_export)
#> Creating new version '20221219T051124Z-a8d18'
#> Writing to pin 'df.0dda107e15c6150535a8d13a54848e37.parquet'
# Before exporting
DBI::dbGetQuery(db_con$con, "SELECT * FROM wflow_tbl;")
#> [1] wflow_id model_tag recipe_tag
#> [4] costs model_type_id py_base_learner_args
#> [7] grid_id df_id random_state
#> <0 rows> (or 0-length row.names)
# After exporting
rpwf_export_db(to_export, db_con)
#> Exporting workflows to db...
#> [1] 1
DBI::dbGetQuery(db_con$con, "SELECT * FROM wflow_tbl;")
#> wflow_id model_tag recipe_tag costs model_type_id
#> 1 1 my_xgboost_model <NA> neg_log_loss 5
#> py_base_learner_args grid_id df_id random_state
#> 1 <NA> 1 1 1004