toad.utils module

class toad.utils.Parallel[source]

Bases: object

apply(func, args=(), kwargs={})[source]
join()[source]
toad.utils.bin_by_splits(feature, splits)[source]

Bin feature by split points

toad.utils.bin_to_number(reg=None)[source]
Returns:func(string) -> number
Return type:function
toad.utils.clip(series, value=None, std=None, quantile=None)[source]

clip series

Parameters:
  • series (array-like) – series need to be clipped
  • value (number | tuple) – min/max value of clipping
  • std (number | tuple) – min/max std of clipping
  • quantile (number | tuple) – min/max quantile of clipping
toad.utils.diff_time(base, target, format=None, time='day')[source]
toad.utils.diff_time_frame(base, frame, format=None)[source]
toad.utils.feature_splits(feature, target)[source]

find posibility spilt points

toad.utils.fillna(feature, by=-1)[source]
toad.utils.generate_str(size=6, chars='ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789')[source]
toad.utils.generate_target(size, rate=0.5, weight=None, reverse=False)[source]

generate target for reject inference

Parameters:
  • size (int) – size of target
  • rate (float) – rate of ‘1’ in target
  • weight (array-like) – weight of ‘1’ to generate target
  • reverse (bool) – if need reverse weight
Returns:

array
toad.utils.get_dummies(dataframe, exclude=None, binary_drop=False, **kwargs)[source]

get dummies

toad.utils.has_nan(arr)[source]
toad.utils.inter_feature(feature, splits)[source]
toad.utils.is_continuous(series)[source]
toad.utils.iter_df(dataframe, feature, target, splits)[source]

iterate dataframe by split points

Returns:iterator (df, splitter)
toad.utils.np_count(arr, value, default=None)[source]
toad.utils.np_unique(arr, **kwargs)[source]
toad.utils.read_json(file)[source]

read json file

toad.utils.save_json(contents, file, indent=4)[source]

save json file

Parameters:
  • contents (dict) – contents to save
  • file (str|IOBase) – file to save
toad.utils.split_target(frame, target)[source]
toad.utils.support_dataframe(require_target=True)[source]

decorator for supporting dataframe

toad.utils.to_ndarray(s, dtype=None)[source]
toad.utils.unpack_tuple(x)[source]