How to use pdpipe - 10 common examples

def test_encode_fit():
    df = _some_df()
    encode_stage = Encode()

    with pytest.raises(UnfittedPipelineStageError):
        encode_stage.transform(df)

    res_df = encode_stage.fit(df)
    assert 'lbl' in res_df.columns
    assert res_df['lbl'][1] == 'acd'
    assert res_df['lbl'][2] == 'alk'
    assert res_df['lbl'][3] == 'alk'

    # see only transform (no fit) when already fitted
    df2 = _some_df2()
    res_df2 = encode_stage.transform(df2)
    assert 'lbl' in res_df2.columns
    assert res_df2['lbl'][1] == 1
    assert res_df2['lbl'][2] == 0

def test_encode_in_pipelin_fit_n_transform():
    drop_name = pdp.ColDrop('name')
    encode_stage = Encode()
    pline = drop_name + encode_stage

    df = _some_df()

    with pytest.raises(UnfittedPipelineStageError):
        res_df = pline.transform(df)

    res_df = pline.fit(df)
    assert 'lbl' in res_df.columns
    assert 'name' in res_df.columns
    assert res_df['lbl'][1] == 'acd'
    assert res_df['lbl'][2] == 'alk'
    assert res_df['lbl'][3] == 'alk'

    res_df = pline.transform(df)
    assert 'lbl' in res_df.columns

def test_encode_in_pipelin_fit_n_transform():
    drop_name = pdp.ColDrop('name')
    encode_stage = Encode()
    pline = drop_name + encode_stage

    df = _some_df()

    with pytest.raises(UnfittedPipelineStageError):
        res_df = pline.transform(df)

    res_df = pline.fit(df)
    assert 'lbl' in res_df.columns
    assert 'name' in res_df.columns
    assert res_df['lbl'][1] == 'acd'
    assert res_df['lbl'][2] == 'alk'
    assert res_df['lbl'][3] == 'alk'

    res_df = pline.transform(df)

def test_row_drop_xor_reducer():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 3]).apply(DF3)
    assert 1 not in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

    res_df = RowDrop([lambda x: x < 3], reduce='xor').apply(DF3)
    assert 1 in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

def test_row_drop_all_reducer():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 3]).apply(DF3)
    assert 1 not in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

    res_df = RowDrop([lambda x: x < 3], reduce='all').apply(DF3)
    assert 1 not in res_df.index
    assert 2 in res_df.index
    assert 3 in res_df.index

def test_row_drop_bad_reducer():
    """Testing the ColDrop pipeline stage."""
    with pytest.raises(ValueError):
        RowDrop([lambda x: x < 2], reduce='al')

def test_row_drop_all_reducer():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 3]).apply(DF3)
    assert 1 not in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

    res_df = RowDrop([lambda x: x < 3], reduce='all').apply(DF3)
    assert 1 not in res_df.index
    assert 2 in res_df.index
    assert 3 in res_df.index

def test_row_drop_columns():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 2]).apply(DF2)
    assert 1 not in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

    res_df = RowDrop([lambda x: x < 2], columns=['a', 'b']).apply(DF2)
    assert 1 not in res_df.index
    assert 2 in res_df.index
    assert 3 in res_df.index

def test_row_drop_xor_reducer():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 3]).apply(DF3)
    assert 1 not in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

    res_df = RowDrop([lambda x: x < 3], reduce='xor').apply(DF3)
    assert 1 in res_df.index
    assert 2 not in res_df.index
    assert 3 in res_df.index

def test_row_drop_verbose():
    """Testing the ColDrop pipeline stage."""
    res_df = RowDrop([lambda x: x < 2]).apply(DF1, verbose=True)
    assert 1 not in res_df.index
    assert 2 in res_df.index
    assert 3 in res_df.index