问题
I've got my pie chart working but I noticed that the text boxes for the actual chart doesn't seem to be working correctly. They are just clustered so I was wondering is there any way for me to move the labels into the middle where the white circle is and have the matching colour beside it or not?
crimeTypes = dict(crimeData["Crime type"].value_counts())
crimeType = []
totalAmount = []
numberOfCrimes = 14
for key in sorted(crimeTypes, key=crimeTypes.get, reverse=True):
crimeType.append(key)
totalAmount.append(crimeTypes.get(key))
crimeType_sample = crimeType[0:numberOfCrimes]
totalAmount_sample = totalAmount[0:numberOfCrimes]
fig1, ax1 = plt.subplots()
ax1.pie(totalAmount_sample, labels=crimeType_sample, autopct='%1.1f%%', shadow=False, startangle=90)
ax1.axis('equal') # Equal aspect ratio ensures that pie is drawn as a circle.
fig1 = plt.gcf()
fig1.set_size_inches(10,10)
circle = plt.Circle(xy=(0,0), radius=0.75, facecolor='white')
plt.gca().add_artist(circle)
plt.show();
回答1:
Here's some sample data to reproduce your problem:
Sample Data:
import pandas as pd
data = (['Burglary']*50 + ['Arson', 'Theft', 'Violence'] + ['Drugs']*10 + ['Other'] +
['Shoplifting']*4 + ['Harassment']*17 + ['Murder', 'Vehicle Crime']*3 +
['Some other Crimes']*12 + ['Even More Crime', 'And Crime', 'And More Crime']*10)
crimeData = pd.DataFrame(data, columns=['Crime type'])
Which will result in this plot:
Use a legend
Do not plot the percentages or labels when you plot, and then create a legend which is placed off to the side:
fig1, ax1 = plt.subplots()
ax1.pie(totalAmount_sample, shadow=False, startangle=90) # No labels or %s
ax1.axis('equal') # Equal aspect ratio ensures that pie is drawn as a circle.
fig1 = plt.gcf()
fig1.set_size_inches(5,5)
circle = plt.Circle(xy=(0,0), radius=0.75, facecolor='white')
plt.gca().add_artist(circle)
plt.legend(labels=[f'{x} {np.round(y/sum(totalAmount_sample)*100,1)}%' for x,y in crimeTypes.items()],
bbox_to_anchor=(1,1))
plt.show();
Rotate the labels:
Create your labels and use rotatelabels=True
. Though this may still appear cramped in many cases.
fig1, ax1 = plt.subplots()
labels=[f'{x} {np.round(y/sum(totalAmount_sample)*100,1)}%' for x,y in crimeTypes.items()]
ax1.pie(totalAmount_sample, labels=labels, shadow=False, startangle=90,
rotatelabels=True) # No %
ax1.axis('equal') # Equal aspect ratio ensures that pie is drawn as a circle.
fig1 = plt.gcf()
fig1.set_size_inches(7,7)
circle = plt.Circle(xy=(0,0), radius=0.75, facecolor='white')
plt.gca().add_artist(circle)
plt.show();
来源:https://stackoverflow.com/questions/53239733/matplotlib-move-labels-into-middle-of-pie-chart