Cmomy

Added

- Now fully support wrapping `xarray.Dataset` objects
- Complete rewrite of wrapper classes
- Added `wrapper` factory methods `cmomy.wrap`, `cmomy.wrap_reduce_vals`, etc.
These automatically select from `CentralMomentsArray` and
`CentralMomentsData`.
- Renamed `CentralMoments` to `CentralMomentsArray` and `xCentralMoments` to
`CentralMomentsData`. The former is parametrized across numpy arrays. The
later is parametrized to work with either `xarray.DataArray` or
`xarray.Dataset` objects.
- Full support `dask` backed `xarray` objects. This will be lazily evaluated.
- Removed `CentralMomentsArray/Data.block` method. Instead, there is a `block`
keyword in `reduce` method.
- `to_dataarray`, `to_dataset`, etc, method of `CentralMomentsArray/Data` now
return a wrapped object instead of an `xarray` object. To access the
underlying `xarray` object, use the `obj` attribute.
- Removed `values` and `data` attributes from `CentralMomentsArray/Data` wrapper
classes. Now wrap either an array or `xarray` object directly (instead of only
really wrapping `numpy` arrays in the old code). To access the wrapped object,
now use `obj` attribute.
- Now include testing of typing (`mypy` and `pyright`) support.

Added

- Added `rolling` module for rolling mean and exponential rolling mean.
- Added `bootstrap_confidence_interval` method to calculate confidence intervals
- Added `moveaxis` function to cleanly handle axes movement of central moments
array
- Added `select_moment` method to select specific moments (weight, average, etc)
from a central moments array
- Added `assign_moment` method to assign values to specific moments.
- Added `vals_to_data` to simplify using `_data` methods for raw values.
- Added `jackknife_` an routines to perform jackknife analysis.
- Update `_vals` routines to properly place the `axis` parameter in result.
- Added `axes_to_end` option to most routines.
- Added `keepdims` option from `reduce_data` and `reduce_vals`

Added

- Added `cmomy.concat` method to concatenate moments objects.

- Added `__getitem__` to `(x)CentralMoments` objects. This method **does not\***
allow changing the moments shape. If you want to do that, you'll need to work
directly with `(x)CentralMoments.to_values()`

Added

- added `cmomy.resample.select_ndat` to select data size along reduction
dimension
- Added `cmomy.randsamp_freq` to top level api

Changed

- Updated `cmomy.resample.randsamp_freq` to select ndat from array

Added

- Added `cmomy.convert.moments_to_comoments` (and
`(x)CentralMoments.moments_to_comoments`)to convert from single variable
moments to comoments. This is useful in `thermoextrap`.
- Added `cmomy.convert.assign_weight` (and `(x)CentralMoments.assign_weights`)
to update weights (useful in `thermoextrap`).

- Added support for `numpy>=2.0.0`. Because we still support older versions, we
still use the old convention for the `copy` parameter to `numpy.array`. Will
change this when minimum numpy is 2.0.

Changed

- Renamed `cmomy.convert` function to `cmomy.convert.moments_type`A bullet item
for the Changed category.

Added

- Now supports python3.12