Box not colored in boxplot

[johngillece]

• altair_recipes version: 0.8.0
• Python version: 3.6.7
• Operating System: MacOS Mohave
• altair version: 4.0.0
• Jupyter Lab

Description

I was looking to generate boxplots with dashes on the top of the the whisker lines and found this tool (thanks!). I generated a boxplot with my own data and noticed that it was only adding a different color to the whisker line but not the box. I subsequently stepped back and attempted to run an example from your website and the same thing happened (image).

What I Did

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import altair as alt
from vega_datasets import data
import altair_recipes as ar
from altair_recipes.common import viz_reg_test
from altair_recipes.display_pweave import show_test

@viz_reg_test
def test_boxplot_color():
    source = data.barley()
    return ar.boxplot(
        source,
        columns=["yield"],
        group_by="year",
        color=True,
        width=800 // len(source["site"].unique()),
    ).facet(column="site")

show_test(test_boxplot_color)

[piccolbo]

My guess this is related to altair 4. What is your expected behavior (besides the discrepancy with the docs which is unquestionable)?

[piccolbo]

My interpretation here is that the fill of the blocks should be colored, the rest black. If you see examples in literature that deviate from this or have other supporting arguments for other colored elements, please let me know. Fix is in master, can you test? Please reopen if anything is wrong.

[johngillece]

It worked. Consider it closed.

import pandas as pd import matplotlib.pyplot as plt import numpy as np import altair as alt from vega_datasets import data from altair import datum import altair_recipes as ar from altair_recipes.common import viz_reg_test from altair_recipes.display_pweave import show_test

Read in the file and fill empty cells with zero

df = pd.read_csv('/Users/jgillece/Desktop/PeriodontalProject/Study2/16S/shannonDivFromExtractionComp/data/metadata.tsv', skiprows=[1], sep='\t')

@viz_reg_test def test_boxplot_color(): source = data.barley() return ar.boxplot( source, columns=["yield"], group_by="year", color=True, width=800 // len(source["site"].unique()), ).facet(column="site")

show_test(test_boxplot_color)

[piccolbo]

I am thinking about waiting for hydrogen to support vega lite 4 before releasing this. Should take < 1 month. Monitoring https://github.com/nteract/hydrogen/issues/1847