def impute(self, mode="last", const_num=0):
'''
Impute the tsdataset by imputing each univariate time series
distinguished by id_col and feature_col
:param mode: imputation mode, select from "last", "const" or "linear".
"last": impute by propagating the last non N/A number to its following N/A.
if there is no non N/A number ahead, 0 is filled instead.
"const": impute by a const value input by user.
"linear": impute by linear interpolation.
:param const_num: indicate the const number to fill, which only effective when mode
is set to "const".
:return: the tsdataset instance.
Note: It is preferred that `impute` is called after `resample` while before
`roll` if needed.
'''
df_list = [impute_timeseries_dataframe(df=self.df[self.df[self.id_col] == id_name],
dt_col=self.dt_col,
mode=mode,
const_num=const_num)
for id_name in self._id_list]
self.df = pd.concat(df_list)
return self
def impute(self, mode="last", const_num=0):
'''
Impute the tsdataset by imputing each univariate time series
distinguished by id_col and feature_col.
:param mode: imputation mode, select from "last", "const" or "linear".
"last": impute by propagating the last non N/A number to its following N/A.
if there is no non N/A number ahead, 0 is filled instead.
"const": impute by a const value input by user.
"linear": impute by linear interpolation.
:param const_num: indicates the const number to fill, which is only effective when mode
is set to "const".
:return: the tsdataset instance.
'''
self.df = self.df.groupby([self.id_col])\
.apply(lambda df: impute_timeseries_dataframe(df=df,
dt_col=self.dt_col,
mode=mode,
const_num=const_num))
self.df.reset_index(drop=True, inplace=True)
return self
def test_impute_timeseries_dataframe(self):
with pytest.raises(AssertionError):
impute_timeseries_dataframe(self.df, dt_col="z")
with pytest.raises(AssertionError):
impute_timeseries_dataframe(
self.df, dt_col="datetime", mode="dummy")
with pytest.raises(AssertionError):
impute_timeseries_dataframe(self.df, dt_col="a")
last_res_df = impute_timeseries_dataframe(
self.df, dt_col="datetime", mode="last")
assert self.df.isna().sum().sum() != 0
assert last_res_df.isna().sum().sum() == 0
const_res_df = impute_timeseries_dataframe(
self.df, dt_col="datetime", mode="const")
assert self.df.isna().sum().sum() != 0
assert const_res_df.isna().sum().sum() == 0
linear_res_df = impute_timeseries_dataframe(
self.df, dt_col="datetime", mode="linear")
assert self.df.isna().sum().sum() != 0
assert linear_res_df.isna().sum().sum() == 0
def impute(self, mode="last", const_num=0):
'''
Impute the tsdataset
:param mode: imputation mode, select from "last", "const" or "linear".
"last": impute by propagating the last non N/A number to its following N/A.
if there is no non N/A number ahead, 0 is filled instead.
"const": impute by a const value input by user.
"linear": impute by linear interpolation.
:param const_num: only effective when mode is set to "const".
'''
df_list = [
impute_timeseries_dataframe(
df=self.df[self.df[self.id_col] == id_name],
dt_col=self.dt_col,
mode=mode,
const_num=const_num) for id_name in self._id_list
]
self.df = pd.concat(df_list)
return self
