Python & Matplotlib : Using Numpy.Array for Stacked Bar Graph

Question

I'm trying to display a graph that includes my data using Stacked Bar Charts. My data is;

new_total = [[5,3,11,17,2,1,5,38,30,45,15,0],[8,21,13,54,21,7,20,103,114,149,77,15],[0,2,6,7,2,6,2,6,22,0,3,0],[0,9,3,11,10,0,0,26,47,17,7,9],[0,11,4,2,5,1,10,35,35,19,16,0],[0,0,0,2,0,0,2,5,6,16,4,3]]

It has 6 elements and every elements represents a color (and every element has 12 subelements). To explain with my codes and pictures;

width = 0.5
ind = np.arange(N)  # the x locations for the groups
temp = []
myMax = 0
myCount = 0
for x in range(len(movies)):
    myCount = myCount + 1
    if myCount == 1:
        self.axes.bar(ind, new_total[0], width)
        temp = np.zeros(N)
    else:
        if x == len(movies) - 1:
            myMax = max(np.add(temp, new_total[x - 1]))
        self.axes.bar(ind, new_total[x], width, bottom=np.add(temp, new_total[x - 1]))

If I'm using this code above; this graph is displayed. As you see for example; purple area is in blue area on somewhere. And total numbers (as you see on left) are wrong.

But if I use this code below;

self.axes.bar(ind, new_total[0], width)
self.axes.bar(ind, new_total[1], width, bottom=np.array(new_total[0]))
self.axes.bar(ind, new_total[2], width, bottom=np.add(new_total[0],new_total[1])) #I could use np.array(new_total[0]) + np.array(new_total[1])
self.axes.bar(ind, new_total[3], width, bottom=np.array(new_total[0]) + np.array(new_total[1]) + np.array(new_total[2]))
self.axes.bar(ind, new_total[4], width, bottom=np.array(new_total[0]) + np.array(new_total[1]) + np.array(new_total[2]) + np.array(new_total[3]))
self.axes.bar(ind, new_total[5], width, bottom=np.array(new_total[0]) + np.array(new_total[1]) + np.array(new_total[2]) + np.array(new_total[3]) + np.array(new_total[4]))

This graph is displayed and thats the graph which I want that shows the colors and total numbers perfectly. But I think this solution is so primitive. Because, sometimes the new_total will have 5 or 7 elements whatever. Can you fix my solution with a perfect way (it can has for loop or whatever)

thats why I'm asking that how can I use for loop for add (... + np.array(new_total[i] ) — Arda Dizdaroğlu
– Arda Dizdaroğlu, Commented Apr 10, 2019 at 14:29

user3053452 · Accepted Answer · 2019-04-10 14:30:30Z

1

I have not tested the code since you have classes and it's not quite minimal working snippet. However, as you can see from your last example, you increase the index by 1 (seems great for for loop). Then you see that the bottom is always the sum of all previous elements, which again goes well with for loop, but here slice notation comes in handy. So given this, the code should look something like this:

for i, sublist in enumerate(new_total):
    self.axes.bar(ind, sublist, bottom=np.sum(new_total[0:i], axis=0), width=width)

A slight caveat is to use, np.sum() function with axis=0 which will sum your array element-wise).

answered Apr 10, 2019 at 14:30

user3053452

6741 gold badge14 silver badges39 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

user3053452 Over a year ago

Let me know how it goes, in case you need any more help.

Arda Dizdaroğlu Over a year ago

so much thanks man, it worked. I have 1 more challenges, if you can help I will be very happy. How can I order the dates in chronological order with their bars?

user3053452 Over a year ago

you'll need to put in a bit more info as I can't see from where you get the dates.

Collectives™ on Stack Overflow

Python & Matplotlib : Using Numpy.Array for Stacked Bar Graph

1 Answer 1

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related