D47crunch

*Released on 2023-07-20*

Command-line interface (CLI)
* Rejoice, you no longer need to know Python: it is now possible to process a multi-session Δ47 dataset with custom anchors and custom UID/sample exclusion list by simply calling `D47crunch -a anchors.csv -e exclude.csv -o outdir rawdata.csv`.

New features
* Add `yspan` option to `D4xdata.plot_residuals()`
* Added `shuffle` option to `virtual_data()`.
* Added `filetype` option to `D4xdata.plot_sessions()`.
* Added `dpi` option to `D4xdata.plot_sessions()`, `D4xdata.plot_residuals()`.

Bugfix
* Fix error in `D4xdata.plot_residuals()` when `hist = False` and `kde = False`.

*Released on 2023-05-16*

Bugfix
* `D4xdata.compute_r()` uses an improved computation for degrees of freedom for arbitrary subsets of sessions and/or samples, yielding more estimates of analytical repeatabilities for Δ47 and Δ48.

New feature
* Added `kde` option to `D4xdata.plot_residuals()`

Other changes
* Minor improvement to y axis tick labels) in `D4xdata.plot_residuals()`.

*Released on 2023-05-14*

New feature
* `D4xdata.plot_bulk_compositions()` plots the dispersion of δ13C and δ18O values for each sample.

*Released on 2023-05-13*

Bugfix
* Eliminate some spurious debugging messages in `_fullcovar()`

*Released on 2023-05-11*

Changes
* Under the hood: constrained parameters in pooled standardization now get fully propagated variance and covariance, allowing for truly arbitrary constraints without having book-keeping problems further down the line.

*Released on 2023-05-11*

Changes
* Graphically improved `D4xdata.plot_distribution_of_analyses()`

Bugfix
* Fix `D4xdata.standardize()` when using weighted sessions