如何影响 colors 列表到 matplotlib 中的直方图索引栏？

Question

I have the the folowing dataframe "freqs2" with index (SD to SD17) and associated values (frequencies):

    freqs
SD  101
SD2 128
...     
SD17 65

I would like to affect a list of precise colors (in order) for each index. I've tried the following code:

colors=['#e5243b','#DDA63A', '#4C9F38','#C5192D','#FF3A21','#26BDE2','#FCC30B','#A21942','#FD6925','#DD1367','#FD9D24','#BF8B2E','#3F7E44','#0A97D9','#56C02B','#00689D','#19486A']
freqs2.plot.bar(freqs2.index, legend=False,rot=45,width=0.85, figsize=(12, 6),fontsize=(14),color=colors )
plt.ylabel('Frequency',fontsize=(17))

As result I obtain all my chart bars in red color (first color of the list).

Based on similar questions, I've tried to integrate "freqs2.index" to stipulate that the list of colors concern index but the problem stay the same.

Answer 1

It looks like a bug in pandas, plotting directly in matplotlib or using seaborn (which I recommend) works:

import seaborn as sns

colors=['#e5243b','#dda63a', '#4C9F38','#C5192D','#FF3A21','#26BDE2','#FCC30B','#A21942','#FD6925','#DD1367','#FD9D24','#BF8B2E','#3F7E44','#0A97D9','#56C02B','#00689D','#19486A']

# # plotting directly with matplotlib works too:
# fig = plt.figure()
# ax = fig.add_axes([0,0,1,1])
# ax.bar(x=df.index, height=df['freqs'], color=colors)

ax = sns.barplot(data=df, x= df.index, y='freqs', palette=colors)
ax.tick_params(axis='x', labelrotation=45)

plt.ylabel('Frequency',fontsize=17)
plt.show()

Edit: an issue already exists on Github