-
-
Notifications
You must be signed in to change notification settings - Fork 8
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
added functionality of step histogram for all three backends #147
added functionality of step histogram for all three backends #147
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #147 +/- ##
==========================================
+ Coverage 76.12% 76.28% +0.16%
==========================================
Files 30 30
Lines 3564 3589 +25
==========================================
+ Hits 2713 2738 +25
Misses 851 851 ☔ View full report in Codecov by Sentry. |
# Ensure alpha and colors have default values | ||
alpha = 1.0 if alpha is unset else alpha | ||
edgecolor = "black" if edgecolor is unset else edgecolor |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
# Ensure alpha and colors have default values | |
alpha = 1.0 if alpha is unset else alpha | |
edgecolor = "black" if edgecolor is unset else edgecolor | |
kwargs = {"bottom": bottom, "fill_color": facecolor, "line_color": edgecolor, "alpha": alpha} |
l_e, | ||
height, | ||
width=widths, | ||
align="edge", | ||
bottom=bottom, | ||
edgecolor=edgecolor, | ||
**_filter_kwargs(kwargs, None, artist_kws), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
l_e, | |
height, | |
width=widths, | |
align="edge", | |
bottom=bottom, | |
edgecolor=edgecolor, | |
**_filter_kwargs(kwargs, None, artist_kws), | |
l_e, height, width=widths, align="edge", **_filter_kwargs(kwargs, None, artist_kws) |
not clear why are you changing
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Actually the bottom and edgecolor attributes are not there in step histogram, so if we feed kwargs (containing bottom and edgecolor values) in to step function then it throws error.
So I removed bottom and edgecolor attributes from kwargs dictionary and I did feed these attributes separately into bar histogram function.
I can create two separate kwargs too for step and bar histograms. Let me know if you want that
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We should then ignore them
**_filter_kwargs({"opacity": alpha}, artist_kws), | ||
) | ||
|
||
if step_hist is False: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
if step_hist
is just True or False ask "if step_hist"
src/arviz_plots/plot_collection.py
Outdated
@@ -935,6 +935,7 @@ def map( | |||
subset_info=False, | |||
store_artist=True, | |||
artist_dims=None, | |||
step_hist=False, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
map is a generic function it should not have arguments specific arguments of specific visuals
src/arviz_plots/plot_collection.py
Outdated
step_hist : boolean, default False | ||
Flag to indicate that the `hist` function should be called with | ||
the keyword argument `step_hist=True` if step histogram is required | ||
instead of bar histogram. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
step_hist : boolean, default False | |
Flag to indicate that the `hist` function should be called with | |
the keyword argument `step_hist=True` if step histogram is required | |
instead of bar histogram. |
src/arviz_plots/plots/ppcdistplot.py
Outdated
@@ -28,6 +28,7 @@ def plot_ppc_dist( | |||
aes_map=None, | |||
plot_kwargs=None, | |||
stats_kwargs=None, | |||
step_hist=False, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
step_hist=False, |
src/arviz_plots/plots/ppcdistplot.py
Outdated
step_hist : boolean, default False | ||
Flag to indicate that the `hist` function should be called with | ||
the keyword argument `step_hist=True` if step histogram is required | ||
instead of bar histogram. | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
step_hist : boolean, default False | |
Flag to indicate that the `hist` function should be called with | |
the keyword argument `step_hist=True` if step histogram is required | |
instead of bar histogram. |
src/arviz_plots/plots/ppcdistplot.py
Outdated
@@ -285,6 +291,7 @@ def plot_ppc_dist( | |||
"observe_density", | |||
data=dt_observed, | |||
ignore_aes=observed_ignore, | |||
step_hist=step_hist, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
step_hist=step_hist, |
src/arviz_plots/visuals/__init__.py
Outdated
@@ -13,7 +13,7 @@ | |||
from arviz_base.labels import BaseLabeller | |||
|
|||
|
|||
def hist(da, target, backend, **kwargs): | |||
def hist(da, target, backend, step_hist=False, **kwargs): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
def hist(da, target, backend, step_hist=False, **kwargs): | |
def hist(da, target, backend,, **kwargs): |
src/arviz_plots/visuals/__init__.py
Outdated
@@ -24,6 +24,7 @@ def hist(da, target, backend, **kwargs): | |||
da.sel(plot_axis="left_edges"), | |||
da.sel(plot_axis="right_edges"), | |||
target, | |||
step_hist=step_hist, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
step_hist=step_hist, |
@aloctavodia all the requested changes have been addressed. Please review. I went with the first method you suggested of passing step value as True in hist key in plot_kwargs parameter for user preference. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I pushed a few changes. Please make sure everything works as expected.
I checked it and everything is working as expected. There were few modifications required to adjust cleaned changes. I have done that. |
great! Can you add a test? |
Should we write a test for step histogram itself, or should we write test for the plots which use kind="hist" ( with step: True in plot_kwargs ) ? @aloctavodia |
the second one |
@aloctavodia I have added the tests for dist_plot ( since that was only using "hist" kind ) Also the docs build is failing on 'pavacalibrationplot.py' file i guess, which i have not touched. Can you check that please? Also review the tests and let me know if more changes are required. Edit: I think you worked today on 'pavacalibrationplot.py' file. So there are some changes which you made which is not here in my code. So doc build is failing because of that reason I guess. |
I made changes to ppcdistplot that are not here, but they should. |
…ave consistent edgecolor
00a75aa
to
c9397a0
Compare
@aloctavodia fixed that. Please review the tests which I added and let me know if anything else needs to be done. |
thanks @The-Broken-Keyboard! |
improve plot_posterior and add tests for it add minimal docstrings add init to test folder update plot_posterior add some docs add docs and implement some feedback improve api docs ignore mystnb warning add plot_trace (arviz-devs#1) * add trace plot * use convert_to_datatree in tests * fix linters * Add to draft API docs --------- Co-authored-by: Oriol Abril-Pla <[email protected]> rename plot_posterior and small improvements fix test rename source file add style add styles update intro plotcollection update intro plotcollection add ecdf to distplot draft trace_dens draft trace_dens draft Update src/arviz_plots/plots/tracedensplot.py Co-authored-by: Oriol Abril-Pla <[email protected]> Update src/arviz_plots/plots/tracedensplot.py Co-authored-by: Oriol Abril-Pla <[email protected]> Update src/arviz_plots/plots/utils.py Co-authored-by: Oriol Abril-Pla <[email protected]> Update src/arviz_plots/plots/utils.py Co-authored-by: Oriol Abril-Pla <[email protected]> fix missing argument try fixing docs fix docs Issue was due to wrong handling of try except. To avoid depending on matplotlib, it is imported within a try except to register the colormaps and styles, but then it is deleted from the namespace which was happening outside the try except and so arviz_plots failed to import on envs without matplotlib rename dens -> dist update api index Add legend method to PlotCollection and improve docs (arviz-devs#32) * start work on legend and improve docs * add anchor placeholder * work on docs * add backend functions to docs and improve plot_collection docs * add missing files * extend api docs * first attempt at bokeh legend * make no-duplication of backend api docs work * run pre-commit * fix and docs on generate_aes_dt * doc improvements * some doc extensions * improve plot_dist docs * fix bokeh legend generation * add warning to legend method Add plot_forest and tests (arviz-devs#34) * some improvements to PlotCollection * add classmethod tests * add more tests * add divergences to plot_trace * extend tests * add divergence data to test_plots * add plot_forest draft For everything to work, the following features were added to: * support for "__variable__" in aesthetics * default values for common aesthetics * support for dict of datatree input -> multiple models aligned within the same plot * add labels to plot_forest still some backend elements missing for bokeh * further improvements to plot_forest * invert y aes to preserve order top-down * improvements to aes_map defaults and shading * attempt support for InferenceData inputs * plot_dist support for multiple models and docs * update dependencies * add more tests * raise error on wrong use of combined * better support for aesthetics in labels elements Add neutral element concept to aes generation (arviz-devs#35) * First try at neutral element * Fix syntax * Try fixing syntax and improve readability * Fix shade aes default * update and extend tests * update aes_dt to aes_dict conversion * run tests with coverage on CI Fixes and improvements to plot_trace_dist (arviz-devs#36) * very rough draft to fix aesthetic mapping defaults * fix plot_trace_dist and add tests * use neutral element for combined elements * add example to plot_trace Extend and rerun tutorials (arviz-devs#38) * add xlabel to plot_trace * rerun plots_intro notebook * update leftover arviz-base references * write use plotcollection notebook and some improvements * fix scale_fig_size defaults * switch docs theme, default to light version PlotCollection coords attribute for better composability (arviz-devs#39) * prepare reuse of plotting functions * call plot_trace within plot_trace_dist * use plot_dist in plot_trace_dist * add contributing section to docs * fix doc warnings * fix dark-light issue with backend logos * write new_plot docs * improve neutral element behaviour and docs * fix a couple cross-references add none backend and hypothesis tests (arviz-devs#41) * add none backend and hypothesis tests * add more tests with hypothesis and scheduled actions * fix hypothesis yaml syntax * use unset as default for none backend so keys are filtered * add branch reference to issue add codecov token (arviz-devs#42) change action to create issue/comment (arviz-devs#44) fix hypothesis action (arviz-devs#46) * add condition to run if testing failed * try fixing branch name * try adding a more specific link to action logs * switch pytest-store_date command order Prepare initial pre-release (arviz-devs#54) * add links between sub-libraries * bump version * use accessor in plot_forest docstring * bump version and configure publishing Fix issue in plot_dist_trace (arviz-devs#64) don't try to rename artists if they haven't been drawn Adding Ridgeplot to Arviz-Plots (arviz-devs#57) * First commit for adding ridgeplot * Updated `plot_ridge` with 'face' and 'edge' artists now and updated visual elements `line_xy` and `fill_between_y` * Modified density calculation pre-check in plot_ridge and modified visual element functions * Added hypothesis tests for plot_ridge * Added ridge_height as a top level arg and other modifications * final changes --------- Co-authored-by: Oriol (ProDesk) <[email protected]> Removed unused 'extra_data' arg (arviz-devs#73) Add plotly backend and gallery prototype (arviz-devs#61) * plotly backend proof of concept * be more consistend with kwarg handling and defaults * restructure backend dependencies and install process * fix typo in bokeh's remove_ticks * complete plotly backend * gallery prototype * extend docs and tests * add plotly equivalences to glossary * more sensible defaults and kwarg handling * automate gallery generation via sphinx extension * figsize and gallery related fixes * full fledged gallery * remove unused references * update gallery generator * initial plotly support for styles * fix gallery generator processing add pp-obs comparison with plot_forest example (arviz-devs#74) * add pp-obs comparison with plot_forest example * improve example expand style functionality (arviz-devs#75) * expand style functionality * add feedback * Apply suggestions from code review Co-authored-by: Oriol Abril-Pla <[email protected]> --------- Co-authored-by: Oriol Abril-Pla <[email protected]> Improve backend documentation and get plot_forest to follow best practices (arviz-devs#78) * start working on best practices and backend docs * use none backend as documentation base * add images for plotly and none * adapt minigallery directive * use none backend instead of arviz_plots.backend * gallery references and plot sizing improvements * fix typo * pylint * wait until using sphinx 8 * pseudo fix for empty minigallery * modify slightly auto sizing * Apply suggestions from code review Co-authored-by: Osvaldo A Martin <[email protected]> * add see also --------- Co-authored-by: Osvaldo A Martin <[email protected]> Histogram support addition to distplot.py (arviz-devs#47) * WIP histogram addition to distplot.py * Added histogram computing if kind='hist' * reformatted histogram dataarrays into a dataset * Modified histogram dataset plot_axis coords to 'x' and 'y' and added visual element function, backend interface and matplotlib backend function for plotting histogram * Allowing xarray_einstats.histogram() function to determine default number of bins * Modified histogram data restructuring function to include bin edge data and modified docstrings for hist backend interface * added plot_dist test parametrizations for kind=kde, hist, ecdf, adjusted hypothesis time limit to 2 seconds and modified backend hist plotting function * added width to histogram plotting, removed print statements from previous commits and updated restructure_hist_data() docstring * switched histogram computation to arviz stats, modified plot_hist visual element slightly for new returned hist density data structure * added 'density=True' to normalize histogram heights and removed axis by default for histograms * added Bokeh backend for hist visual element and removed ecdf parametrization from test_plot_dist_models * updated docstring, added 'alpha' argument to the 'hist' backend plotting functions, renamed 'plot_hist' to 'hist', modified `remove_axis` logic slightly and set density=True as default in stats_kwargs * Added 'hist' to visuals.rst * deactivate tests for hist kind and multiple models It needs a fix in arviz-stats to work * removed restructure_dist and glyph default artist kwarg in bokeh * updated hist backend interface and matplotlib hist backend function with updates fromrootogram plot * plotly hist and plot_dist improvements --------- Co-authored-by: Oriol Abril-Pla <[email protected]> Add plot_compare (arviz-devs#77) * add plot_compare * directly use plot_backend * add new kwargs * use fill_between_y * docs * remove commented code * use plot_kwargs * use plotcollection * use plotcollection * alow disabling elements * pass pc_kwargs to plotcollection * try to fix example in gallery * add missing import * Update gallery_generator.py * Improve show method for plotcollection * fix 1x1 grid generation in plotly * fix plotly 1x1 plots * add basic test * fix tests * isort * remove redundant array conversion --------- Co-authored-by: Oriol Abril-Pla <[email protected]> rework styles (arviz-devs#88) * rework color palletes * update plotly clean template fix link to github (arviz-devs#91) Add ESS Plot (arviz-devs#58) * First commit for essplot and scatter_xy visual element * update for ess plot and addition of 'x' aesthetic for 'model' dim * addition of quantile plot and updated x aesthetic mapping * Added rugplot to essplot * updates to essplot * fixed default value for arg 'extra_methods' * modified scatter_xy visual element to take into account _process_da_x_y update * Added color/linestyles aesthetics and simplified min_ess plotting * Added annotate_xy visual element, applied to essplot for extra_methods * visual element vertical alignment logic modification and arviz-stats compute_ranks addition attempt * added docs for essplot * tests for essplot * added scatter_xy to visuals.rst * added rug=True to example gallery plot_ess_local * fixes for rugplot issue and hypothesis test failures * shifted mean_ess, sd_ess computing to before plot_kwargs check+artist plotting logic and modified hypothesis tests * Updated plot_ess and tests * updated .toml file for arviz-stats dependency * Modified figsize to more of a plot_forest approach, fixed order of plots in plots.rst and expanded max limit for test methods in testplots.py in .pylintrc * Updated plot_ess docstring * Switched from .grid to .wrap, removed unused figsize coeffs, disabled pylint warning on testplots.py * final fixes now only waiting for us to figure out behaviour and scope in arviz-stats and xarray-einstats * update pyproject requirements * pylint --------- Co-authored-by: Oriol (ProDesk) <[email protected]> fix warning (arviz-devs#99) Adding Plot ESS Evolution (arviz-devs#71) * Initial ess evolution plot updated plot_ess_evolution including a common ess_dataset computing func added mean and sd annotations like essplot docs and example gallery for plot_ess_evolution updated verticalalign logic for mean/sd and correct (although overlaid and not flattened yet) rug is now displayed removed rug plot added tests updated scatter_xy func to plot_ess version fixed docstring altered store_artist for xlabel, ylabel and modified hypothesis tests shifted mean_ess, sd_edd computing to before plot_kwargs check+artist plotting logic updated docstring, added figsizing and set vertical_align for mean and sd text kwargs as setdefault removed 'rankdata' branch of arviz-stats from dependencies docstring typo fix gallery-generator updated for documentation building * remove visual duplicated when rebasing --------- Co-authored-by: Oriol (ProDesk) <[email protected]> Add plot_psense_dist (arviz-devs#93) * Add plot_psense_dist * Update src/arviz_plots/plots/psensedistplot.py Co-authored-by: Oriol Abril-Pla <[email protected]> * Update src/arviz_plots/plots/psensedistplot.py Co-authored-by: Oriol Abril-Pla <[email protected]> * concat da and simplify logic * set sample_dims to sample * refactor * minor fixes and update pyproject to install from GH * support sample_dims argument and all backends * add minigallery to docstring * tweak example * ensure pointinterval only plot does not have yticks * add initial test for psense plot * add test and example * fix test * fix docstring * rename __group__ --------- Co-authored-by: Oriol Abril-Pla <[email protected]> pin datatree and prepare release (arviz-devs#103) * pin datatree and prepare release * update naming for hist_dim to match arviz-stats install arviz-base/stats from github (arviz-devs#104) move out new_ds to arviz-stats (arviz-devs#102) Bump codecov/codecov-action from 4 to 5 (arviz-devs#107) Bumps [codecov/codecov-action](https://github.com/codecov/codecov-action) from 4 to 5. - [Release notes](https://github.com/codecov/codecov-action/releases) - [Changelog](https://github.com/codecov/codecov-action/blob/main/CHANGELOG.md) - [Commits](codecov/codecov-action@v4...v5) --- updated-dependencies: - dependency-name: codecov/codecov-action dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <[email protected]> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> update version, dependencies and CI (arviz-devs#110) * update version, dependencies and CI * skip 3.13 until numba release * update rtd env * ignore line too long in gallery template string Use DataTree class from xarray (arviz-devs#111) * start work for xarray datatree compatibility * use datatree from xarray * fix docs * remove unused import Update pyproject.toml (arviz-devs#113) Add energy plot (arviz-devs#108) * add energy plot * remove __variable__, add example * hardcode sampler_dims * use legend only with matplotlib Add plot for distribution of convergence diagnostics (arviz-devs#105) * Add plot for distribution of convergence diagnostics * add ref_line and methods for r-hat * fix docstring * rename, add example * update gallery example * use rhat instead of rhat_rank * update sphinx.configuration key * update sphinx.configuration key * add vline/hline. Co-authored-by: Oriol Abril-Pla <[email protected]> * fix test * fix docstring * fix docstring --------- Co-authored-by: Oriol Abril-Pla <[email protected]> Add separated prior and likelihood groups (arviz-devs#117) * Add separated prior and likelihood groups * upper bound for plotly Add psense_quantities plot (arviz-devs#119) * add psense_quantities plot * remove comment * fix conflicting dimension * update docts * fix color * fix ls * split quantities * fix docstring rename arviz-clean to arviz-variat (arviz-devs#120) rename arviz-clean to arviz-variat add cetrino and vibrant styles to plotly (arviz-devs#121) psense: fix facetting and add xlabel (arviz-devs#123) * fix facetting * add x-label Add summary dictionary arguments (arviz-devs#125) * add summary dictionary arguments * fix spelling plotly: change format of title update in backend (arviz-devs#124) * plotly: change format of title update in backend * update the pyproject.toml file to restrict lower bound of plotly version to 6 Update glossary.md (arviz-devs#126) Add PAV-adjusted calibration plot (arviz-devs#127) * draft pava ppc * add pava-adjusted calibration plot * use dt * fix var name upper bound plotly (arviz-devs#128) use isotonic function that work with datatrees (arviz-devs#131) add reference, fix xlabel (arviz-devs#132) Fix bug when setting some plot_kwargs to false (arviz-devs#134) * fix bug setting when setting some plot_kwargs to false * remove references Add citations (arviz-devs#135) * add citations * reformat citations * fix indentation * fix links * add reference file use <6 version of plotly for documentation and use latest for other purposes (arviz-devs#136) * temporary fix for plotly-plot-rendering-on-webpage * include plotly also in readthedocs fix see algo pava gallery (arviz-devs#137) Add plot_ppc_dist (arviz-devs#138) * add plot_ppc_dist * remove comments * add test and small fixes * fix typo Add warning message for discrete data (arviz-devs#139) * add warning message for discrete data * do not fail on warnings rename plot_pava and minor fixes (arviz-devs#140) fix excesive margins (arviz-devs#141) add arviz-style for bokeh (arviz-devs#122) * add arviz-styles for bokeh * add arviz-styles for bokeh * Update arviz-variat.yml * add more styles Add rootogram (arviz-devs#142) fix examples (arviz-devs#144) Reorganize categories in the gallery (arviz-devs#145) * reorganize categories gallery * update condig * rename remove plots from titles (arviz-devs#146) * remove plots from titles * more renaming * more renaming consistence data_pairs, remove markers pava (arviz-devs#152) added functionality of step histogram for all three backends (arviz-devs#147) * added functionality of step histogram for all three backends * changed user preference format for step histograms * clean * little modifications to make it compatible with cleaned code and to have consistent edgecolor * added test for plot_dist with step value as true * minor fixes: step hist test is not required for 'none' backend --------- Co-authored-by: aloctavodia <[email protected]> use continuous outcome for plot_ppc_dist example (arviz-devs#154) add grid visual (arviz-devs#155) all test cases are passed ruff changes seperating plotting functionality of bayes_factor testing get_plotting funcitonality new implementation of arviz-plot Update bfplot.py it should be something like this. Please, check that it works properly, add docstrings and improve when necessary Improved the plot_bf() function, applied fixes, and added additional assert for test cases Fixed the check for ref_val Added a docstring for plot_bf() Change the backend from matplotlib to none Added a legend to componenet_groups Added few more asserts in test cases Did some other minor fixing and refractoring Signed-off-by: PiyushPanwarFST <[email protected]> Enhance plot_bf() Visualization & Documentation, Fix Linter Issues Adjusted and re-centered BF values for better graph representation. Added plot_bf() function to the documentation. Resolved pylint linter errors in code. Signed-off-by: PiyushPanwarFST <[email protected]> deleting some files and modifying some files
This implements the new functionality of step histograms discussed in #96
step
function available in matplotlib backendstep
function available in bokeh backendgo.scatter
functionality to implement step histogram functionalitystep_hist flag parameter is added at appropriate places, with default value as False, to maintain compatiblity with old version of functionalities related to histogram.
How to choose step style for histograms :
User will be able to choose step histograms by passing True value to step key of hist key in plot_kwargs parameter, as shown below:
plot_kwargs = {"hist": {"step": True}}
Here are examples using plot_dist function in arviz_plots with flag step_hist as True :

Backend: Plotly
Backend: Matplotlib

Backend: Bokeh

📚 Documentation preview 📚: https://arviz-plots--147.org.readthedocs.build/en/147/