How to get rid of decimal point in the legend-CodePudding

Here is the code I used to make the visualization:

results = pd.DataFrame(reduced_data,columns=['pca1','pca2','pca3','pca4','pca5','pca6','pca7'])
ax = sns.scatterplot(x="pca1", y="pca2", hue=H12022_full['cluster'],palette=['green','orange','brown','dodgerblue', 'yellow', 'blue', 'red'], data=results)
ax.get_xaxis().set_major_formatter(mpl.ticker.StrMethodFormatter('{x:,.0f}'))
ax.get_yaxis().set_major_formatter(mpl.ticker.StrMethodFormatter('{x:,.0f}'))

This is the result:

I wanted to know how I could get rid of the decimals in the legend. So instead of 1.0 it will be just 1?

CodePudding user response：

You can try this. As the values are in float (1.0, 2.0..) and in text, you will need to convert it to float and then integer... A simple example.

df = pd.DataFrame({'foo':[1,2,3.0,4,5], 'bar': [1,2,3.2,4,5]})
ax = sns.scatterplot(x='foo', y='bar', data=df, hue='foo', palette=['green','orange','brown','dodgerblue', 'yellow'])
l = ax.get_legend_handles_labels()[1]
legend = list(map(round, list(map(float, l))))
plt.legend(legend)

Plot

CodePudding user response：

Can you try the following:

actual_legends = ax.get_legend_handles_labels()[1]
legend = [int(float(l)) for l in actual_legends]
plt.legend(legend)