Creating a barchart from histogram, python matplotlib

Question

I have an histogram like this:

Where my data are stored with an append in that way:

(while parsing the file)
{
 [...]
 a.append(int(number))
 #a = [1,1,2,1,1, ]... 
}

plt.hist(a, 180)

But as you can see from the image, there are lot of blank areas, so I would like to build a barchart from this data, how can I reorganize them like:

#a = [ 1: 4023, 2: 3043, 3:...]

Where 1 is the "number" and 4023 is an example on how many "hit" of the number 1? From what I have seen in this way I can call:

plt.bar(...)

and creating it, so that I can show only the relevant numbers, with more readability. If there is a simple way to cut white area in the Histo is also welcome.

I would like also to show the top counter of each columns, but I have no idea how to do it.

Answer 1

Assuming you have some numpy array a full of integers then the code below will produce the bar chart you desire.

It uses np.bincount to count the number of values, note that it only works for non-negative integers.

Also note that I have adjusted the indices so that the plot centrally rather than to the left (using ind-width/2.).

import matplotlib.pyplot as plt
import numpy as np

# Generate some random data.
N=300
a = np.random.random_integers(low=0, high=20, size=N)

# Use bincount and nonzero to generate your data in the correct format.
b = np.bincount(a)
ind = np.nonzero(b)[0]

width=0.8

fig, ax = plt.subplots()

ax.bar(ind-width/2., b)

plt.show()