arviz_plots.plot_rank

Contents

arviz_plots.plot_rank#

arviz_plots.plot_rank(dt, *, var_names=None, filter_vars=None, group='posterior', coords=None, sample_dims=None, envelope_prob=None, method='mtc_c', thin=None, plot_collection=None, backend=None, labeller=None, aes_by_visuals=None, visuals=None, stats=None, **pc_kwargs)[source]#

Fractional rank Δ-ECDF plots.

Rank plots are built by replacing the posterior draws by their ranking computed over all chains. Then each chain is plotted independently. If all of the chains are targeting the same posterior, we expect the ranks in each chain to be uniformly distributed. To simplify comparison we compute the ordered fractional ranks, which are distributed uniformly in [0, 1]. Additionally, we plot the Δ-ECDF, that is, the difference between the expected CDF from the observed ECDF.

The points that contribute the most to deviations from uniformity are computed as described in [1].

Parameters:
dtxarray.DataTree

Input data

var_namesstr or list of str, optional

One or more variables to be plotted. Currently only one variable is supported. Prefix the variables by ~ when you want to exclude them from the plot.

filter_vars{None, “like”, “regex”}, optional, default=None

If None (default), interpret var_names as the real variables names. If “like”, interpret var_names as substrings of the real variables names. If “regex”, interpret var_names as regular expressions on the real variables names.

groupstr, optional

Which group to use. Defaults to “posterior”.

coordsdict, optional

Coordinates to plot.

sample_dimsstr or sequence of hashable, optional

Dimensions to reduce unless mapped to an aesthetic. Defaults to rcParams["data.sample_dims"]

envelope_probfloat, optional

Indicates the probability that should be contained within the envelope. Defaults to rcParams["stats.envelope_prob"].

method{“mtc_c”, “envelope”}, default “mtc_c”

Method to use for the rank plot. If “mtc_c”, the multi-chain test is performed and suspicious points are highlighted. If “envelope”, the envelope is computed and plotted.

thinbool, default None

Whether to thin the data before plotting. Defaults to None, which means that it is set to True if “method” is “envelope” and False otherwise.

plot_collectionPlotCollection, optional
backend{“matplotlib”, “bokeh”, “plotly”}, optional
labellerlabeller, optional
aes_by_visualsmapping of {strsequence of str}, optional

Mapping of visuals to aesthetics that should use their mapping in plot_collection when plotted. Valid keys are the same as for visuals.

visualsmapping of {strmapping or bool}, optional

Valid keys are:

statsmapping, optional

Valid keys are:

  • ecdf_pit -> passed to ecdf_pit. Default is {"n_simulations": 1000}.

  • mtc_c -> passed to mchain_uniformity_test.

  • thin -> passed to thin

**pc_kwargs

Passed to arviz_plots.PlotCollection.wrap

Returns:
PlotCollection

Notes

The preferred method is mtc_c as it takes into account the autocorrelation in the rank values as described in [1]. The “envelope” method is not longer recommended and it will likely be removed in a future release.

References

[1] (1,2)

Tasso et al. LOO-PIT predictive model checking arXiv:2603.02928 (2026).

[2]

Säilynoja et al. Graphical test for discrete uniformity and its applications in goodness-of-fit evaluation and multiple sample comparison. Statistics and Computing 32(32). (2022) https://doi.org/10.1007/s11222-022-10090-6

Examples

Rank plot for the crabs hurdle-negative-binomial dataset.

>>> from arviz_plots import plot_rank, style
>>> style.use("arviz-variat")
>>> from arviz_base import load_arviz_data
>>> dt = load_arviz_data('crabs_hurdle_nb')
>>> plot_rank(dt, var_names=["~mu"])
../../_images/arviz_plots-plot_rank-1.png

faceted plot with fractional ranks for each variable

Rank plot

Rank plot
Contents