Skip to content

indexing module

Classes and functions for indexing.

build_param_indexer function

build_param_indexer(
    param_names,
    class_name='ParamIndexer',
    module_name=None
)

A factory to create a class with parameter indexing.

Parameter indexer enables accessing a group of rows and columns by a parameter array (similar to loc). This way, one can query index/columns by another Series called a parameter mapper, which is just a pd.Series that maps columns (its index) to params (its values).

Parameter indexing is important, since querying by column/index labels alone is not always the best option. For example, pandas doesn't let you query by list at a specific index/column level.

Args

param_names : list of str
Names of the parameters.
class_name : str
Name of the generated class.
module_name : str
Name of the module to which the class should be bound.

Usage

>>> from vectorbtpro import *
>>> from vectorbtpro.base.indexing import build_param_indexer, indexing_on_mapper

>>> MyParamIndexer = build_param_indexer(['my_param'])
>>> class C(MyParamIndexer):
...     def __init__(self, df, param_mapper):
...         self.df = df
...         self._my_param_mapper = param_mapper
...         super().__init__([param_mapper])
...
...     def indexing_func(self, pd_indexing_func):
...         return type(self)(
...             pd_indexing_func(self.df),
...             indexing_on_mapper(self._my_param_mapper, self.df, pd_indexing_func)
...         )

>>> df = pd.DataFrame({'a': [1, 2], 'b': [3, 4]})
>>> param_mapper = pd.Series(['First', 'Second'], index=['a', 'b'])
>>> c = C(df, param_mapper)

>>> c.my_param_loc['First'].df
0    1
1    2
Name: a, dtype: int64

>>> c.my_param_loc['Second'].df
0    3
1    4
Name: b, dtype: int64

>>> c.my_param_loc[['First', 'First', 'Second', 'Second']].df
      a     b
0  1  1  3  3
1  2  2  4  4

get_idxs function

get_idxs(
    idxr,
    index=None,
    columns=None,
    freq=None,
    template_context=None,
    **kwargs
)

Translate indexer to row and column indices.

If idxr is not an indexer class, wraps it with Idxr.

Keyword arguments are passed when constructing a new Idxr.

get_index_points function

get_index_points(
    index,
    every=None,
    normalize_every=False,
    at_time=None,
    start=None,
    end=None,
    exact_start=False,
    on=None,
    add_delta=None,
    kind=None,
    indexer_method='bfill',
    indexer_tolerance=None,
    skip_not_found=True
)

Translate indices or labels into index points.

See PointIdxr for arguments.

Usage

  • Provide nothing to generate at the beginning:
>>> from vectorbtpro import *

>>> index = pd.date_range("2020-01", "2020-02", freq="1d")

>>> vbt.get_index_points(index)
array([0])
  • Provide every as an integer frequency to generate index points using NumPy:
>>> # Generate a point every five rows
>>> vbt.get_index_points(index, every=5)
array([ 0,  5, 10, 15, 20, 25, 30])

>>> # Generate a point every five rows starting at 6th row
>>> vbt.get_index_points(index, every=5, start=5)
array([ 5, 10, 15, 20, 25, 30])

>>> # Generate a point every five rows from 6th to 16th row
>>> vbt.get_index_points(index, every=5, start=5, end=15)
array([ 5, 10])
  • Provide every as a time delta frequency to generate index points using Pandas:
>>> # Generate a point every week
>>> vbt.get_index_points(index, every="W")
array([ 4, 11, 18, 25])

>>> # Generate a point every second day of the week
>>> vbt.get_index_points(index, every="W", add_delta="2d")
array([ 6, 13, 20, 27])

>>> # Generate a point every week, starting at 11th row
>>> vbt.get_index_points(index, every="W", start=10)
array([11, 18, 25])

>>> # Generate a point every week, starting exactly at 11th row
>>> vbt.get_index_points(index, every="W", start=10, exact_start=True)
array([10, 11, 18, 25])

>>> # Generate a point every week, starting at 2020-01-10
>>> vbt.get_index_points(index, every="W", start="2020-01-10")
array([11, 18, 25])
  • Instead of using every, provide indices explicitly:
>>> # Generate one point
>>> vbt.get_index_points(index, on="2020-01-07")
array([6])

>>> # Generate multiple points
>>> vbt.get_index_points(index, on=["2020-01-07", "2020-01-14"])
array([ 6, 13])

get_index_ranges function

get_index_ranges(
    index,
    index_freq=None,
    every=None,
    normalize_every=False,
    split_every=True,
    start_time=None,
    end_time=None,
    lookback_period=None,
    start=None,
    end=None,
    exact_start=False,
    fixed_start=False,
    closed_start=True,
    closed_end=False,
    add_start_delta=None,
    add_end_delta=None,
    kind=None,
    skip_not_found=True,
    jitted=None
)

Translate indices, labels, or bounds into index ranges.

See RangeIdxr for arguments.

Usage

  • Provide nothing to generate one largest index range:
>>> from vectorbtpro import *

>>> index = pd.date_range("2020-01", "2020-02", freq="1d")

>>> np.column_stack(vbt.get_index_ranges(index))
array([[ 0, 32]])
  • Provide every as an integer frequency to generate index ranges using NumPy:
>>> # Generate a range every five rows
>>> np.column_stack(vbt.get_index_ranges(index, every=5))
array([[ 0,  5],
       [ 5, 10],
       [10, 15],
       [15, 20],
       [20, 25],
       [25, 30]])

>>> # Generate a range every five rows, starting at 6th row
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every=5,
...     start=5
... ))
array([[ 5, 10],
       [10, 15],
       [15, 20],
       [20, 25],
       [25, 30]])

>>> # Generate a range every five rows from 6th to 16th row
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every=5,
...     start=5,
...     end=15
... ))
array([[ 5, 10],
       [10, 15]])
  • Provide every as a time delta frequency to generate index ranges using Pandas:
>>> # Generate a range every week
>>> np.column_stack(vbt.get_index_ranges(index, every="W"))
array([[ 4, 11],
       [11, 18],
       [18, 25]])

>>> # Generate a range every second day of the week
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     add_start_delta="2d"
... ))
array([[ 6, 11],
       [13, 18],
       [20, 25]])

>>> # Generate a range every week, starting at 11th row
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     start=10
... ))
array([[11, 18],
       [18, 25]])

>>> # Generate a range every week, starting exactly at 11th row
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     start=10,
...     exact_start=True
... ))
array([[10, 11],
       [11, 18],
       [18, 25]])

>>> # Generate a range every week, starting at 2020-01-10
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     start="2020-01-10"
... ))
array([[11, 18],
       [18, 25]])

>>> # Generate a range every week, each starting at 2020-01-10
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     start="2020-01-10",
...     fixed_start=True
... ))
array([[11, 18],
       [11, 25]])

>>> # Generate an expanding range that increments by week
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     start=0,
...     exact_start=True,
...     fixed_start=True
... ))
array([[ 0,  4],
       [ 0, 11],
       [ 0, 18],
       [ 0, 25]])
  • Use a look-back period (instead of an end index):
>>> # Generate a range every week, looking 5 days back
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     lookback_period=5
... ))
array([[ 6, 11],
       [13, 18],
       [20, 25]])

>>> # Generate a range every week, looking 2 weeks back
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     every="W",
...     lookback_period="2W"
... ))
array([[ 0, 11],
       [ 4, 18],
       [11, 25]])
  • Instead of using every, provide start and end indices explicitly:
>>> # Generate one range
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     start="2020-01-01",
...     end="2020-01-07"
... ))
array([[0, 6]])

>>> # Generate ranges between multiple dates
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     start=["2020-01-01", "2020-01-07"],
...     end=["2020-01-07", "2020-01-14"]
... ))
array([[ 0,  6],
       [ 6, 13]])

>>> # Generate ranges with a fixed start
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     start="2020-01-01",
...     end=["2020-01-07", "2020-01-14"]
... ))
array([[ 0,  6],
       [ 0, 13]])
  • Use closed_start and closed_end to exclude any of the bounds:
>>> # Generate ranges between multiple dates
>>> # by excluding the start date and including the end date
>>> np.column_stack(vbt.get_index_ranges(
...     index,
...     start=["2020-01-01", "2020-01-07"],
...     end=["2020-01-07", "2020-01-14"],
...     closed_start=False,
...     closed_end=True
... ))
array([[ 1,  7],
       [ 7, 14]])

indexing_on_mapper function

indexing_on_mapper(
    mapper,
    ref_obj,
    pd_indexing_func
)

Broadcast mapper Series to ref_obj and perform pandas indexing using pd_indexing_func.

normalize_idxs function

normalize_idxs(
    idxs,
    target_len
)

Normalize indexes into a 1-dim integer array.

AutoIdxr class

AutoIdxr(
    *args,
    **kwargs
)

Class for resolving indices, datetime-like objects, frequency-like objects, and labels for one axis.

Superclasses

Inherited members

above_to_len field

Whether to place len(index) instead of -1 if AutoIdxr.value is above the last index.

below_to_zero field

Whether to place 0 instead of -1 if AutoIdxr.value is below the first index.

closed_end field

Whether slice end should be inclusive.

closed_start field

Whether slice start should be inclusive.

idxr_kwargs field

Keyword arguments passed to the selected indexer.

indexer_method field

Method for pd.Index.get_indexer.

kind field

Kind of value.

Allowed are

If None, will (try to) determine automatically based on the type of indices.

level field

One or more levels.

If level is not None and kind is None, kind becomes "labels".

value field

One or more integer indices, datetime-like objects, frequency-like objects, or labels.

Can also be an instance of PosSel holding position(s) and LabelSel holding label(s).

ColIdxr class

ColIdxr(
    idxr,
    **idxr_kwargs
)

Class for resolving column indices.

Superclasses

Inherited members

idxr field

Indexer.

Can be an instance of UniIdxr, a custom template, or a value to be wrapped with AutoIdxr.

idxr_kwargs field

Keyword arguments passed to AutoIdxr.

DTCIdxr class

DTCIdxr(
    *args,
    **kwargs
)

Class for resolving indices provided as datetime-like components.

Superclasses

Inherited members

closed_end field

Whether slice end should be inclusive.

closed_start field

Whether slice start should be inclusive.

get_dtc_namedtuple static method

DTCIdxr.get_dtc_namedtuple(
    value=None,
    **parse_kwargs
)

Convert a value to a DTCNT instance.

jitted field

Jitting option passed to index_matches_dtc_nb and index_within_dtc_range_nb.

parse_kwargs field

Keyword arguments passed to DTC.parse.

value field

One or more datetime-like components.

DatetimeIdxr class

DatetimeIdxr(
    *args,
    **kwargs
)

Class for resolving indices provided as datetime-like objects.

Superclasses

Inherited members

above_to_len field

Whether to place len(index) instead of -1 if DatetimeIdxr.value is above the last index.

below_to_zero field

Whether to place 0 instead of -1 if DatetimeIdxr.value is below the first index.

closed_end field

Whether slice end should be inclusive.

closed_start field

Whether slice start should be inclusive.

indexer_method field

Method for pd.Index.get_indexer.

Allows two additional values: "before" and "after".

value field

One or more datetime-like objects.

ExtPandasIndexer class

ExtPandasIndexer(
    **kwargs
)

Extension of PandasIndexer that also implements indexing using xLoc.

Superclasses

Inherited members

Subclasses

xloc class property

Subclass of iLoc that transforms an Idxr-based operation with get_idxs to an iLoc operation.

IdxDict class

IdxDict(
    *args,
    **kwargs
)

Class for building an index setter from a dict.

Superclasses

Inherited members

index_dct field

Dict that contains indexer objects as keys and values to be set as values.

IdxFrame class

IdxFrame(
    *args,
    **kwargs
)

Class for building an index setter from a DataFrame.

Superclasses

Inherited members

colidx_kwargs field

Keyword arguments passed to colidx if the indexer isn't an instance of ColIdxr.

df field

DataFrame or any array-like object to create the DataFrame from.

rowidx_kwargs field

Keyword arguments passed to rowidx if the indexer isn't an instance of RowIdxr.

split field

Whether to split the setting operation.

If False, will set all values using a single operation. Otherwise, the following options are supported:

  • 'columns': one operation per column
  • 'rows': one operation per row
  • True or 'elements': one operation per element

IdxRecords class

IdxRecords(
    *args,
    **kwargs
)

Class for building index setters from records - one per field.

Superclasses

Inherited members

col_field field

Column field.

If None or True, will search for "col", "column", and "symbol" (case-insensitive).

If a record doesn't have a column field, all columns will be set. If there's no row and column field, the field value will become the default of the entire array.

colidx_kwargs field

Keyword arguments passed to colidx if the indexer isn't an instance of ColIdxr.

records field

Series, DataFrame, or any sequence of mapping-like objects.

If a Series or DataFrame and the index is not a default range, the index will become a row field. If a custom row field is provided, the index will be ignored.

row_field field

Row field.

If None or True, will search for "row", "index", "open time", and "date" (case-insensitive). If IdxRecords.records is a Series or DataFrame, will also include the index name if the index is not a default range.

If a record doesn't have a row field, all rows will be set. If there's no row and column field, the field value will become the default of the entire array.

rowidx_kwargs field

Keyword arguments passed to rowidx if the indexer isn't an instance of RowIdxr.

IdxSeries class

IdxSeries(
    *args,
    **kwargs
)

Class for building an index setter from a Series.

Superclasses

Inherited members

idx_kwargs field

Keyword arguments passed to idx if the indexer isn't an instance of Idxr.

split field

Whether to split the setting operation.

If False, will set all values using a single operation. Otherwise, will do one operation per element.

sr field

Series or any array-like object to create the Series from.

IdxSetter class

IdxSetter(
    *args,
    **kwargs
)

Class for setting values based on indexing.

Superclasses

Inherited members

fill_and_set method

IdxSetter.fill_and_set(
    shape,
    keep_flex=False,
    fill_value=nan,
    **kwargs
)

Fill a new array and set its values based on IdxSetter.get_set_meta.

If keep_flex is True, will return the most memory-efficient array representation capable of flexible indexing.

If fill_value is None, will search for the _def key in IdxSetter.idx_items. If there's none, will be set to NaN.

get_set_meta method

IdxSetter.get_set_meta(
    shape,
    index=None,
    columns=None,
    freq=None,
    template_context=None
)

Get meta of setting operations in IdxSetter.idx_items.

idx_items field

Items where the first element is an indexer and the second element is a value to be set.

set method

IdxSetter.set(
    arr,
    set_funcs=None,
    **kwargs
)

Set values of a NumPy array based on IdxSetter.get_set_meta.

set_col_idxs class method

IdxSetter.set_col_idxs(
    arr,
    idxs,
    v
)

Set column indices in an array.

set_pd method

IdxSetter.set_pd(
    pd_arr,
    **kwargs
)

Set values of a Pandas array based on IdxSetter.get_set_meta.

set_row_and_col_idxs class method

IdxSetter.set_row_and_col_idxs(
    arr,
    row_idxs,
    col_idxs,
    v
)

Set row and column indices in an array.

set_row_idxs class method

IdxSetter.set_row_idxs(
    arr,
    idxs,
    v
)

Set row indices in an array.

IdxSetterFactory class

IdxSetterFactory()

Class for building index setters.

Superclasses

Inherited members

Subclasses

get method

IdxSetterFactory.get()

Get an instance of IdxSetter or a dict of such instances - one per array name.

Idxr class

Idxr(
    *idxrs,
    **idxr_kwargs
)

Class for resolving indices.

Superclasses

Inherited members

idxr_kwargs field

Keyword arguments passed to RowIdxr and ColIdxr.

idxrs field

A tuple of one or more indexers.

If one indexer is provided, can be an instance of RowIdxr or ColIdxr, a custom template, or a value to wrapped with RowIdxr.

If two indexers are provided, can be an instance of RowIdxr and ColIdxr respectively, or a value to wrapped with RowIdxr and ColIdxr respectively.

IdxrBase class

IdxrBase()

Abstract class for resolving indices.

Superclasses

Inherited members

Subclasses

check_idxs method

IdxrBase.check_idxs(
    idxs,
    check_minus_one=False
)

Check indices after resolving them.

get method

IdxrBase.get(
    *args,
    **kwargs
)

Get indices.

slice_indexer class method

IdxrBase.slice_indexer(
    index,
    slice_,
    closed_start=True,
    closed_end=False
)

Compute the slice indexer for input labels and step.

IndexingBase class

IndexingBase()

Class that supports indexing through IndexingBase.indexing_func.

Superclasses

Inherited members

Subclasses

  • PandasIndexer
  • vectorbtpro.indicators.custom.adx.ParamIndexer
  • vectorbtpro.indicators.custom.atr.ParamIndexer
  • vectorbtpro.indicators.custom.bbands.ParamIndexer
  • vectorbtpro.indicators.custom.hurst.ParamIndexer
  • vectorbtpro.indicators.custom.ma.ParamIndexer
  • vectorbtpro.indicators.custom.macd.ParamIndexer
  • vectorbtpro.indicators.custom.msd.ParamIndexer
  • vectorbtpro.indicators.custom.obv.ParamIndexer
  • vectorbtpro.indicators.custom.ols.ParamIndexer
  • vectorbtpro.indicators.custom.patsim.ParamIndexer
  • vectorbtpro.indicators.custom.pivotinfo.ParamIndexer
  • vectorbtpro.indicators.custom.rsi.ParamIndexer
  • vectorbtpro.indicators.custom.sigdet.ParamIndexer
  • vectorbtpro.indicators.custom.stoch.ParamIndexer
  • vectorbtpro.indicators.custom.supertrend.ParamIndexer
  • vectorbtpro.indicators.custom.vwap.ParamIndexer
  • vectorbtpro.labels.generators.bolb.ParamIndexer
  • vectorbtpro.labels.generators.fixlb.ParamIndexer
  • vectorbtpro.labels.generators.fmax.ParamIndexer
  • vectorbtpro.labels.generators.fmean.ParamIndexer
  • vectorbtpro.labels.generators.fmin.ParamIndexer
  • vectorbtpro.labels.generators.fstd.ParamIndexer
  • vectorbtpro.labels.generators.meanlb.ParamIndexer
  • vectorbtpro.labels.generators.pivotlb.ParamIndexer
  • vectorbtpro.labels.generators.trendlb.ParamIndexer
  • vectorbtpro.signals.generators.ohlcstx.ParamIndexer
  • vectorbtpro.signals.generators.ohlcstx.ParamIndexer
  • vectorbtpro.signals.generators.rand.ParamIndexer
  • vectorbtpro.signals.generators.randnx.ParamIndexer
  • vectorbtpro.signals.generators.randx.ParamIndexer
  • vectorbtpro.signals.generators.rprob.ParamIndexer
  • vectorbtpro.signals.generators.rprobnx.ParamIndexer
  • vectorbtpro.signals.generators.rprobx.ParamIndexer
  • vectorbtpro.signals.generators.rprobx.ParamIndexer
  • vectorbtpro.signals.generators.stx.ParamIndexer
  • vectorbtpro.signals.generators.stx.ParamIndexer

indexing_func method

IndexingBase.indexing_func(
    pd_indexing_func,
    **kwargs
)

Apply pd_indexing_func on all pandas objects in question and return a new instance of the class.

Should be overridden.

indexing_setter_func method

IndexingBase.indexing_setter_func(
    pd_indexing_setter_func,
    **kwargs
)

Apply pd_indexing_setter_func on all pandas objects in question.

Should be overridden.

IndexingError class

IndexingError(
    *args,
    **kwargs
)

Exception raised when an indexing error has occurred.

Superclasses

  • builtins.BaseException
  • builtins.Exception

LabelIdxr class

LabelIdxr(
    *args,
    **kwargs
)

Class for resolving indices provided as labels.

Superclasses

Inherited members

closed_end field

Whether slice end should be inclusive.

closed_start field

Whether slice start should be inclusive.

level field

One or more levels.

value field

One or more labels.

Loc class

Loc(
    indexing_func,
    indexing_setter_func=None,
    **kwargs
)

Forwards pd.Series.loc/pd.DataFrame.loc operation to each Series/DataFrame and returns a new class instance.

Superclasses

Inherited members

LocBase class

LocBase(
    indexing_func,
    indexing_setter_func=None,
    **kwargs
)

Class that implements location-based indexing.

Superclasses

Inherited members

Subclasses

indexing_func class property

Indexing function.

indexing_kwargs class property

Keyword arguments passed to LocBase.indexing_func.

indexing_setter_func class property

Indexing setter function.

MaskIdxr class

MaskIdxr(
    *args,
    **kwargs
)

Class for resolving indices provided as a mask.

Superclasses

Inherited members

value field

Mask.

PandasIndexer class

PandasIndexer(
    **kwargs
)

Implements indexing using iloc, loc, xs and __getitem__.

Usage

>>> from vectorbtpro import *
>>> from vectorbtpro.base.indexing import PandasIndexer

>>> class C(PandasIndexer):
...     def __init__(self, df1, df2):
...         self.df1 = df1
...         self.df2 = df2
...         super().__init__()
...
...     def indexing_func(self, pd_indexing_func):
...         return type(self)(
...             pd_indexing_func(self.df1),
...             pd_indexing_func(self.df2)
...         )

>>> df1 = pd.DataFrame({'a': [1, 2], 'b': [3, 4]})
>>> df2 = pd.DataFrame({'a': [5, 6], 'b': [7, 8]})
>>> c = C(df1, df2)

>>> c.iloc[:, 0]
<__main__.C object at 0x1a1cacbbe0>

>>> c.iloc[:, 0].df1
0    1
1    2
Name: a, dtype: int64

>>> c.iloc[:, 0].df2
0    5
1    6
Name: a, dtype: int64

Superclasses

Inherited members

Subclasses

iloc class property

Forwards pd.Series.iloc/pd.DataFrame.iloc operation to each Series/DataFrame and returns a new class instance.

indexing_kwargs class property

Indexing keyword arguments.

loc class property

Forwards pd.Series.loc/pd.DataFrame.loc operation to each Series/DataFrame and returns a new class instance.

xs method

PandasIndexer.xs(
    *args,
    **kwargs
)

Forwards pd.Series.xs/pd.DataFrame.xs operation to each Series/DataFrame and returns a new class instance.

ParamLoc class

ParamLoc(
    mapper,
    indexing_func,
    indexing_setter_func=None,
    level_name=None,
    **kwargs
)

Access a group of columns by parameter using pd.Series.loc.

Uses mapper to establish link between columns and parameter values.

Superclasses

Inherited members

encode_key class method

ParamLoc.encode_key(
    key
)

Encode key.

get_idxs method

ParamLoc.get_idxs(
    key
)

Get array of indices affected by this key.

level_name class property

Level name.

mapper class property

Mapper.

PointIdxr class

PointIdxr(
    *args,
    **kwargs
)

Class for resolving index points.

Superclasses

Inherited members

add_delta field

Offset to be added to each in on.

Gets converted to a proper offset/timedelta using to_freq.

at_time field

Time of the day either as a (human-readable) string or datetime.time.

Every datetime in on gets floored to the daily frequency, while at_time gets converted into a timedelta using time_to_timedelta and added to add_delta. Index must be datetime-like.

end field

End index/date.

If (human-readable) string, gets converted into a datetime.

If every is None, gets used to filter the final index array.

every field

Frequency either as an integer or timedelta.

Gets translated into on array by creating a range. If integer, an index sequence from start to end (exclusive) is created and 'indices' as kind is used. If timedelta-like, a date sequence from start to end (inclusive) is created and 'labels' as kind is used.

If at_time is not None and every and on are None, every defaults to one day.

exact_start field

Whether the first index should be exactly start.

Depending on every, the first index picked by pd.date_range may happen after start. In such a case, start gets injected before the first index generated by pd.date_range.

indexer_method field

Method for pd.Index.get_indexer.

Allows two additional values: "before" and "after".

indexer_tolerance field

Tolerance for pd.Index.get_indexer.

If at_time is set and indexer_method is neither exact nor nearest, indexer_tolerance becomes such that the next element must be within the current day.

kind field

Kind of data in on: indices or labels.

If None, gets assigned to indices if on contains integer data, otherwise to labels.

If kind is 'labels', on gets converted into indices using pd.Index.get_indexer. Prior to this, gets its timezone aligned to the timezone of the index. If kind is 'indices', on gets wrapped with NumPy.

normalize_every field

Normalize start/end dates to midnight before generating date range.

on field

Index/label or a sequence of such.

Gets converted into datetime format whenever possible.

skip_not_found field

Whether to drop indices that are -1 (not found).

start field

Start index/date.

If (human-readable) string, gets converted into a datetime.

If every is None, gets used to filter the final index array.

PosIdxr class

PosIdxr(
    *args,
    **kwargs
)

Class for resolving indices provided as integer positions.

Superclasses

Inherited members

value field

One or more integer positions.

RangeIdxr class

RangeIdxr(
    *args,
    **kwargs
)

Class for resolving index ranges.

Superclasses

Inherited members

add_end_delta field

Offset to be added to each in end.

If string, gets converted to a proper offset/timedelta using to_freq.

add_start_delta field

Offset to be added to each in start.

If string, gets converted to a proper offset/timedelta using to_freq.

closed_end field

Whether end should be inclusive.

closed_start field

Whether start should be inclusive.

end field

End index/label or a sequence of such.

Gets converted into datetime format whenever possible.

Gets broadcasted together with start.

end_time field

End time of the day either as a (human-readable) string or datetime.time.

Every datetime in end gets floored to the daily frequency, while end_time gets converted into a timedelta using time_to_timedelta and added to add_end_delta. Index must be datetime-like.

every field

Frequency either as an integer or timedelta.

Gets translated into start and end arrays by creating a range. If integer, an index sequence from start to end (exclusive) is created and 'indices' as kind is used. If timedelta-like, a date sequence from start to end (inclusive) is created and 'bounds' as kind is used.

If start_time and end_time are not None and every, start, and end are None, every defaults to one day.

exact_start field

Whether the first index in the start array should be exactly start.

Depending on every, the first index picked by pd.date_range may happen after start. In such a case, start gets injected before the first index generated by pd.date_range.

Cannot be used together with lookback_period.

fixed_start field

Whether all indices in the start array should be exactly start.

Works only together with every.

Cannot be used together with lookback_period.

jitted field

Jitting option passed to Resampler.map_bounds_to_source_ranges.

kind field

Kind of data in on: indices, labels or bounds.

If None, gets assigned to indices if start and end contain integer data, to bounds if start, end, and index are datetime-like, otherwise to labels.

If kind is 'labels', start and end get converted into indices using pd.Index.get_indexer. Prior to this, get their timezone aligned to the timezone of the index. If kind is 'indices', start and end get wrapped with NumPy. If kind` is 'bounds', Resampler.map_bounds_to_source_ranges is used.

lookback_period field

Lookback period either as an integer or offset.

If lookback_period is set, start becomes end-lookback_period. If every is not None, the sequence is generated from start+lookback_period to end and then assigned to end.

If string, gets converted to a proper offset/timedelta using to_freq. If integer, gets multiplied by the frequency of the index if the index is not integer.

normalize_every field

Normalize start/end dates to midnight before generating date range.

skip_not_found field

Whether to drop indices that are -1 (not found).

split_every field

Whether to split the sequence generated using every into start and end arrays.

After creation, and if split_every is True, an index range is created from each pair of elements in the generated sequence. Otherwise, the entire sequence is assigned to start and end, and only time and delta instructions can be used to further differentiate between them.

Forced to False if every, start_time, and end_time are not None and fixed_start is False.

start field

Start index/label or a sequence of such.

Gets converted into datetime format whenever possible.

Gets broadcasted together with end.

start_time field

Start time of the day either as a (human-readable) string or datetime.time.

Every datetime in start gets floored to the daily frequency, while start_time gets converted into a timedelta using time_to_timedelta and added to add_start_delta. Index must be datetime-like.

RowIdxr class

RowIdxr(
    idxr,
    **idxr_kwargs
)

Class for resolving row indices.

Superclasses

Inherited members

idxr field

Indexer.

Can be an instance of UniIdxr, a custom template, or a value to be wrapped with AutoIdxr.

idxr_kwargs field

Keyword arguments passed to AutoIdxr.

UniIdxr class

UniIdxr()

Abstract class for resolving indices based on a single index.

Superclasses

Inherited members

Subclasses

UniIdxrOp class

UniIdxrOp(
    op_func,
    *idxrs
)

Class for applying an operation to one or more indexers.

Produces a single set of indices.

Superclasses

Inherited members

idxrs field

A tuple of one or more indexers.

op_func field

Operation function that takes the indices of each indexer (as *args), index (keyword argument), and freq (keyword argument), and returns new indices.

hslice class

hslice(
    start=_Missing.MISSING,
    stop=_Missing.MISSING,
    step=_Missing.MISSING
)

Hashable slice.

Superclasses

Inherited members

from_slice class method

hslice.from_slice(
    slice_
)

Construct from a slice.

start field

Start.

step field

Step.

stop field

Stop.

to_slice method

hslice.to_slice()

Convert to a slice.

iLoc class

iLoc(
    indexing_func,
    indexing_setter_func=None,
    **kwargs
)

Forwards pd.Series.iloc/pd.DataFrame.iloc operation to each Series/DataFrame and returns a new class instance.

Superclasses

Inherited members

Subclasses

index_dict class

index_dict(
    *args,
    **kwargs
)

Dict that contains indexer objects as keys and values to be set as values.

Each indexer object must be hashable. To make a slice hashable, use hslice. To make an array hashable, convert it into a tuple.

To set a default value, use the _def key (case-sensitive!).

Superclasses

Inherited members

pdLoc class

pdLoc(
    indexing_func,
    indexing_setter_func=None,
    **kwargs
)

Forwards a Pandas-like indexing operation to each Series/DataFrame and returns a new class instance.

Superclasses

Inherited members

Subclasses

pd_indexing_func class method

pdLoc.pd_indexing_func(
    obj,
    key
)

Pandas-like indexing operation.

pd_indexing_setter_func class method

pdLoc.pd_indexing_setter_func(
    obj,
    key,
    value
)

Pandas-like indexing setter operation.

xLoc class

xLoc(
    indexing_func,
    indexing_setter_func=None,
    **kwargs
)

Subclass of iLoc that transforms an Idxr-based operation with get_idxs to an iLoc operation.

Superclasses

Inherited members