import numpy as np
import pandas as pd
from .. import keywords as EATK
from ..timeseries.create.from_intervals import flatten_and_fill
# =============================================================================
# flatten overconsumption
# =============================================================================
[docs]
def test_flatten_empty():
"""Empty dataframe is flattened as empty"""
assert flatten_and_fill(pd.DataFrame()).empty
[docs]
def test_flatten_one_interval():
"""Check that flattening is OK in one row table."""
start = pd.Timestamp("2020-02-29")
end = start + pd.Timedelta("1min")
test_data = pd.DataFrame(
[[start, end, 42.0]], columns=[EATK.start_f, EATK.end_f, "test"],
)
flat_consumption = flatten_and_fill(
test_data, start_f=EATK.start_f, end_f=EATK.end_f, time_f=EATK.time_f,
)
expect_index = pd.DatetimeIndex([start, end], name=EATK.time_f)
expect_series = pd.DataFrame(
np.array([42.0, np.nan]).reshape(2, 1),
index=expect_index,
columns=["test"],
)
pd.testing.assert_frame_equal(
expect_series, flat_consumption, check_dtype=False, check_freq=False,
)
[docs]
def test_flatten():
"""Check that flattening looks OK."""
time_f = "timestamp"
end_f = "end_custom"
start_f = "start_custom"
begin = pd.Timestamp("2022-12-05 18:00:00")
starts = pd.date_range(start=begin, freq="10min", periods=3, name=start_f)
targets = pd.date_range(start=begin, freq="5min", periods=6, name=time_f)
data = {
start_f: starts,
end_f: starts + pd.Timedelta("5min"),
"col_str": ["toto"] * 3,
"col_float": np.arange(0.0, 6.0, 2.0),
}
table = pd.DataFrame.from_dict(data)
# Default filling
flat_expect = pd.DataFrame.from_dict(
{
"col_str": ["toto"] * 6,
"col_float": np.arange(0, 6),
},
)
flat_expect.index = targets
flat_expect.iloc[[1, 3, 5], :] = np.nan
flat_table = flatten_and_fill(table, end_f=end_f, start_f=start_f)
pd.testing.assert_frame_equal(flat_expect, flat_table, check_freq=False)
# explicit filling
flat_filled_table = flatten_and_fill(
table, fill_values={"col_str": "tata"}, end_f=end_f, start_f=start_f,
)
flat_expect["col_str"] = ["toto", "tata"] * 3
pd.testing.assert_frame_equal(
flat_expect,
flat_filled_table,
check_freq=False,
check_dtype=False,
)
