I created a scatter plot using matplotlib but I am somehow unable to get the labels to center into the boxes within the colorbar..

This is the code I have so far:

cMap = ListedColormap(['Orange', 'Purple', 'Blue','Red','Green'])
fig, ax = plt.subplots()
plt.figure(figsize=(12,12),dpi = 80)

#data
dist = np.random.rand(1900,1900)

#legend
cbar = plt.colorbar(scatter)

cbar.ax.get_yaxis().set_ticks([])
for j, lab in enumerate(['$Training$','$None$','$GS$','$ML$','$Both$']):
    cbar.ax.text( .5, j - .985, lab, ha='left', va='center', rotation = 270)
cbar.ax.get_yaxis().labelpad = 15
cbar.ax.set_ylabel('Outliers', rotation=270)

indices = np.where(outlier_label != -2)[0]
plt.scatter(dist[indices, 0], dist[indices, 1], c=outlier_label[indices], cmap=cMap, s=20)
plt.gca().set_aspect('equal', 'datalim')
plt.title('Projection of the data', fontsize=24)

Thanks!

CodePudding user response：

In line cbar.ax.text( .5, j - .985, lab, ha='left', va='center', rotation = 270) you have to work and change with '.985' with try and error to get better results.

CodePudding user response：

You can extract the y limits of the colorbar to know its top and bottom. Dividing that area into 11 equally spaced positions, will have the 5 centers at the odd positions of that list. Similarly, you can extract the x limits to find the horizontal center.

Some remarks:

• If you already called plt.subplots(), then plt.figure() will create a new figure, leaving the first plot empty. You can set the figsize directly via plt.subplots(figsize=...)
• You are mixing matplotlib's