solarized dark custom style sheet

[gennaro-tedesco]

I would like to propose a custom style sheet in line with the solarized dark colour palette. The style mirrors the ones already present and is the only addition to this pull request.

[story645]

Hi can you please provide an example of the new stylesheet and explain why it's worth adopting?

[timhoffm]

The solarized palette comes in light and dark flavor.

We currently have a solarized light style but not a dark correspondent. It makes sense to add the dark style as well.

My concern is that the dark and light style should be the same except for the colors. Currently, there are differences with:

• figure.frameon
• figure.edgecolor
• axes.grid
• grid.color

Not sure why the light style is called "Solarize_Light2". That's odd in multiple ways:

• All other styles are lowercase
• It's called "solarize" whereas the theme is "solarized"
• No idea why 2. At least now, we don't have another solarize style.

[timhoffm]

Please use spaces, not tabs.

[gennaro-tedesco]

done.

[gennaro-tedesco]

My concern is that the dark and light style should be the same except for the colors. Currently, there are differences with:

I could mirror the current light theme if you believe this is the standard we have to follow, only changing the corresponding colour cycles and palettes. I could wait for your decision on the topic and make the changes accordingly.

[gennaro-tedesco]

Moreover, find below a small collection of examples of what this theme looks like, obtained by using seaborn functionalities as shown here or here

import matplotlib
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
plt.style.use('solarized_dark')

x = np.random.normal(0, 1, size=30)
sns.kdeplot(x)
sns.kdeplot(x, bw=.2, label="bw: 0.2")
sns.kdeplot(x, bw=2, label="bw: 2")
plt.legend()
plt.show()

tips = sns.load_dataset("tips")
sns.catplot(x="sex", y="total_bill",
                hue="smoker", col="time",
                data=tips, kind="bar",
                height=4, aspect=.7)
plt.show()

import pandas as pd 
index = pd.date_range("1 1 2000", periods=100,
                    freq="m", name="date")

data = np.random.randn(100, 4).cumsum(axis=0)
wide_df = pd.DataFrame(data, index, ["a", "b", "c", "d"])
sns.lineplot(data=wide_df)
plt.show()

[QuLogic]

I think if we have light, we should also have a dark variant, but they should be consistent in what they change.

I never understood the naming of our light Solarized theme, and no-one answered when I asked. I think we should add a new one with more consistent name (and whatever consistencies we need with a dark variant) and (soft?-)deprecate the old one.

[gennaro-tedesco]

I never understood the naming of our light Solarized theme, and no-one answered when I asked. I think we should add a new one with more consistent name (and whatever consistencies we need with a dark variant) and (soft?-)deprecate the old one.

If you think the current proposal for the dark theme satisfies the requirements I could create the equivalent for the light version, so that the two mirror each other except for the actual colour palettes.

[jklymak]

This fell off everyone's radar, but I think would be a useful addition...