nb module¶

Numba-compiled functions for resampling.

date_range_nb function¶

date_range_nb(
    start,
    end,
    freq=numpy.timedelta64(86400000000000,'ns'),
    incl_left=True,
    incl_right=True
)

Generate a datetime index with nanosecond precision from a date range.

Inspired by pandas.date_range.

index_difference_nb function¶

index_difference_nb(
    source_index,
    target_index
)

Get the elements in source_index not present in target_index.

last_before_target_index_nb function¶

last_before_target_index_nb(
    source_index,
    target_index,
    incl_source=True,
    incl_target=False
)

For each source index, find the position of the last source index between the original source index and the corresponding target index.

map_bounds_to_source_ranges_nb function¶

map_bounds_to_source_ranges_nb(
    source_index,
    target_lbound_index,
    target_rbound_index,
    closed_lbound=True,
    closed_rbound=False,
    skip_not_found=False
)

Get the source bounds that correspond to the target bounds.

Returns a 2-dim array where the first column is the absolute start index (including) nad the second column is the absolute end index (excluding).

If an element cannot be mapped, the start and end of the range becomes -1.

map_index_to_source_ranges_nb function¶

map_index_to_source_ranges_nb(
    source_index,
    target_index,
    target_freq=None,
    before=False
)

Get the source bounds that correspond to each target index.

If target_freq is not None, the right bound is limited by the frequency in target_freq. Otherwise, the right bound is the next index in target_index.

Returns a 2-dim array where the first column is the absolute start index (including) and the second column is the absolute end index (excluding).

If an element cannot be mapped, the start and end of the range becomes -1.

map_to_target_index_nb function¶

map_to_target_index_nb(
    source_index,
    target_index,
    target_freq=None,
    before=False,
    raise_missing=True
)

Get the index of each from source_index in target_index.

If before is True, applied on elements that come before and including that index. Otherwise, applied on elements that come after and including that index.

If raise_missing is True, will throw an error if an index cannot be mapped. Otherwise, the element for that index becomes -1.

resample_source_mask_nb function¶

resample_source_mask_nb(
    source_mask,
    source_index,
    target_index,
    source_freq=None,
    target_freq=None
)

Resample a source mask to the target index.

Becomes True only if the target bar is fully contained in the source bar. The source bar is represented by a non-interrupting sequence of True values in the source mask.