def test_compress_series(series, before, expected, inference):
assert series.dtype == before
compressed_series = compress_func(
series,
typeset=StandardSet(),
compressor=SparseCompressor(),
with_inference=inference,
inplace=False,
)
assert compressed_series.dtype == expected
def test_copy_frame():
df = pd.DataFrame({"column": pd.Series([1], dtype="int64")})
compressed_df = compress_func(
df,
typeset=StandardSet(),
compressor=DefaultCompressor(),
with_inference=True,
inplace=False,
)
assert id(df) != id(compressed_df)
def test_compress_series(series, before, expected):
assert series.dtype == before
compressed_series = compress_func(
series,
typeset=StandardSet(),
compressor=DefaultCompressor(),
with_inference=True,
inplace=False,
)
assert str(compressed_series.dtype) == expected
assert_series_equal(series, compressed_series, check_dtype=False)
def _(
data: pd.Series,
typeset: VisionsTypeset,
compressor: BaseTypeCompressor,
with_inference: bool,
units: str = "megabytes",
) -> None:
before = data.dtype
compressed = compress_func(data, typeset, compressor, with_inference)
after = compressed.dtype
if str(before) != str(after):
print(
f'{data.name}: converting from {before} to {after} saves {savings(data, compressed, units)} (use `data[{data.name}].astype("{after}")`)'
)