Reputation: 41
I was asked to make a bar chart using this data:
customer=['Alice', 'Bob', 'Claire']
cakes=[5,9,7]
flavor=['chocolate', 'vanilla', 'strawberry']
The end result should appear like this.
The code to create the chart using Altair is as follows:
import altair
data = altair.Data(customer=['Alice', 'Bob', 'Claire'], cakes=[5,9,7], flavor=['chocolate', 'vanilla', 'strawberry'])
chart = altair.Chart(data)
mark = chart.mark_bar()
enc = mark.encode(x='customer:N',y='cakes',color='flavor:N')
enc.display()
The code I used to generate a similar chart using matplotlib is as follows:
import matplotlib.pyplot as plt
customers = ['Alice', 'Bob', 'Clair']
length = [0,1,2]
cakes = [5, 9, 7]
flavors = ['chocolate', 'vanilla', 'strawberry']
colors = ['brown', 'beige', 'magenta']
for w,x,y,z in zip(length, flavors, colors, cakes):
plt.bar(w, z, color = y, align = 'center', alpha = 0.5, label = x)
plt.xticks(y_pos, customers)
plt.ylabel('Cakes')
plt.title('Customers')
plt.legend()
plt.show()
My question is: How may I more elegantly assign multiple labels and colors to a legend?
Please also give me suggestions as to how I may design this code so that it is more elegant and concise, thank you and sorry for such a long question!
Upvotes: 0
Views: 1785
Reputation: 150735
Since you are customize too many things, e.g. color, name, and especially legend label for each bar, I don't see any significantly better solution. One slight improvement is to remove length/y_pos
:
for w,x,y,z in zip(customers, flavors, colors, cakes):
plt.bar(w, z, color = y, align = 'center', alpha = 0.5, label = x)
plt.ylabel('Cakes')
plt.title('Customers')
plt.legend()
plt.show()
Output:
Upvotes: 2