def test_save():
filepath = "./datasets/configs/params.json"
params = {
"param1": 69,
"param2": {
"nestedParam": 420
},
"Split": 6969,
}
save_config(filepath, params)
contents = read_config(filepath)
assert params == contents
def test_save():
filepath = "./datasets/configs/params.json"
params = {
"param1": 69,
"param2": {"nestedParam": 420},
"Split": 6969,
"df": "./datasets/encoding/test.csv",
}
save_config(filepath, params)
contents = read_config(filepath)
params.pop("df")
assert params == contents
def test_config():
df = pd.DataFrame({"A": np.arange(1, 100), "B": np.arange(1, 100)})
_ = df.to_csv("./datasets/configs/dataset.csv", index=False)
params = {
"col_1": "A",
"col_2": "B",
"test_size": 0.2,
}
config_path = "./datasets/configs/pipeline_config.json"
save_config(config_path, params)
pipeline = BasePipeline(
train_df_path="./datasets/configs/dataset.csv",
steps=[times_two, squared, split],
config_file=config_path,
custom_reader=custom_read,
)
pipeline.process()
assert len(pipeline.params["X_train"]) == 80
pipeline.remove("split")
pipeline.process()
assert (pipeline.params["train_df"].shape[0] ==
pipeline.params["train_df_copy"].shape[0])