def test_add(self):
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",
"test_size": 0.2,
}
pipeline = Pipeline(
df_path="./datasets/configs/dataset.csv",
steps=[times_two, split],
params=params,
)
pipeline.process()
assert pipeline.params["df"].loc[42, "A"] == df.loc[42, "A"] * 2
pipeline.add(
squared,
{
"col_2": "A",
},
before="times_two",
)
pipeline.process()
num_0 = pipeline.params["df"].loc[42, "A"]
num_1 = df.loc[42, "A"]
assert num_0 == (num_1**2) * 2
pipeline.remove("squared")
pipeline.add(squared, {"col_2": "A"}, after="read_file")
pipeline.process()
num_0 = pipeline.params["df"].loc[42, "A"]
num_1 = df.loc[42, "A"]
assert num_0 == (num_1**2) * 2
def test_remove(self):
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,
}
pipeline = Pipeline(
df_path="./datasets/configs/dataset.csv",
steps=[times_two, squared, split],
params=params,
)
pipeline.process()
assert len(pipeline.params["X_train"]) == 80
pipeline.remove("split")
pipeline.process()
assert pipeline.params["df"].shape[0] == df.shape[0]
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 = {
"df": "./datasets/configs/dataset.csv",
"col_1": "A",
"col_2": "B",
"test_size": 0.2,
}
config_path = "./datasets/configs/pipeline_config.json"
save_config(config_path, params)
pipeline = Pipeline(
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["df"].shape[0] == pipeline.params["df_copy"].shape[0])