Apply hash pattern to BAR PLOT

Question

I would like to create a list where i can add in English speaking countries and make sure that reflects on my barplot

top9 = master_frame[master_frame['country_code'].isin(["USA","CHN","GBR","IND","CAN","ISR","FRA","DEU","SWE"])]

plt.figure(figsize=(10,5))
ax2=sns.barplot(x='country_code', y='raised_amount_usd', data=top9, estimator=np.sum,)
ax2.set_yscale('log')
ax2.set(xlabel='Funding by country', ylabel='Total amount of investments')
ax2.set_title('Investment distribution Country wise',fontsize =18, weight='bold')
plt.show()

Output:

Here the challenge is that i need to highlight countries which have English as an official language and so my thought was to add a hash pattern to the bars of the countries and show a hue box above.

Any idea how i can achieve this??

Thank you in advance!

Answer 1

I am not quite sure about the hash pattern you mentioned.

Here is my understanding, you want to achieve two things.

First, color the bars having different property.

Second, create a legend box on the plot.

The approach below is way to fulfill your purpose.

---

1. Color the bars.

This step can be easily done with setting up the parameter palette in the function sns.barplot. You shall input a sequence of colors into it in hex format (there also lots of other ways exist).

color_eng = "#94698b"
color_non_eng = "#4369ef"

# the sequence length is the number of bars.
palette = [color_eng, color_eng, color_non_eng, color_eng, color_non_eng]

ax2 = sns.barplot(x='country_code', y='raised_amount_usd', 
                  data=top9, estimator=np.sum, palette=palette)

2. Create a legend.

To create a customized legend box, we could set up the handles of the axes with manually created patches (I borrowed the code idea from this SO answer).

import matplotlib.patches as mpatches

p_eng = mpatches.Patch(color=color_eng, label='English speaking country')
p_non_eng = mpatches.Patch(color=color_non_eng, label='non-English speaking country')

ax2.legend(handles=[p_eng, p_non_eng])

---

The final code shall looks like:

import matplotlib.patches as mpatches

top9 = master_frame[master_frame['country_code'].isin(["USA","CHN","GBR","IND","CAN","ISR","FRA","DEU","SWE"])]

plt.figure(figsize=(10,5))

color_eng = "#94698b"
color_non_eng = "#4369ef"

# the sequence length is the number of bars.
# Shall be composed by you
palette = [color_eng, color_eng, color_non_eng, color_eng, .......]

ax2 = sns.barplot(x='country_code', y='raised_amount_usd', 
                  data=top9, estimator=np.sum, palette=palette)

p_eng = mpatches.Patch(color=color_eng, label='English speaking country')
p_non_eng = mpatches.Patch(color=color_non_eng, label='non-English speaking country')

ax2.legend(handles=[p_eng, p_non_eng])

ax2.set_yscale('log')
ax2.set(xlabel='Funding by country', ylabel='Total amount of investments')
ax2.set_title('Investment distribution Country wise',fontsize =18, weight='bold')
plt.show()