user3580316 user3580316 - 4 months ago 17
Python Question

Frequency diagram with matplotlib

I am trying to automate a frequency diagram with matplotlib in Python in order to count occurences, instead of having to manually plot in Excel. However, I am not able to make an similar as possible diagram as I have done in Excel. Is this possible with Matplotlib?

In Excel:

enter image description here

Code:

#!/usr/bin/python

import numpy as np
import matplotlib.pyplot as plt
from numpy import *
import os
import sys
import csv
from random import randint

x = [6,0,0,26,0,0,0,0,5,0,7,0,12,12,0,0,0,3,0,5,5,0,10,4,3,5,1,0,2,0,0,1,0,8,0,3,7,1,0,0,0,1,1,0,0,0,0,0,7,16,0,0,0,5]


plt.hist(x)
plt.title("Frequency diagram")
plt.xlabel("Value")
plt.ylabel("Frequency")
plt.show()


Result (The readability is not as good compared to Excel, how can I make it as similar as the excel graph):

enter image description here

Answer
import numpy as np
import matplotlib.pyplot as plt

def make_hist(ax, x, bins=None, binlabels=None, width=0.85, extra_x=1, extra_y=4, 
              text_offset=0.3, title=r"Frequency diagram", 
              xlabel="Values", ylabel="Frequency"):
    if bins is None:
        xmax = max(x)+extra_x
        bins = range(xmax+1)
    if binlabels is None:
        if np.issubdtype(np.asarray(x).dtype, np.integer):
            binlabels = [str(bins[i]) if bins[i+1]-bins[i] == 1 else 
                         '{}-{}'.format(bins[i], bins[i+1]-1)
                         for i in range(len(bins)-1)]
        else:
            binlabels = [str(bins[i]) if bins[i+1]-bins[i] == 1 else 
                         '{}-{}'.format(*bins[i:i+2])
                         for i in range(len(bins)-1)]
        if bins[-1] == np.inf:
            binlabels[-1] = '{}+'.format(bins[-2])
    n, bins = np.histogram(x, bins=bins)
    patches = ax.bar(range(len(n)), n, align='center', width=width)
    ymax = max(n)+extra_y

    ax.set_xticks(range(len(binlabels)))
    ax.set_xticklabels(binlabels)

    ax.set_title(title)
    ax.set_xlabel(xlabel)
    ax.set_ylabel(ylabel)
    ax.set_ylim(0, ymax)
    ax.grid(True, axis='y')
    # http://stackoverflow.com/a/28720127/190597 (peeol)
    ax.spines['top'].set_visible(False)
    ax.spines['right'].set_visible(False)
    ax.spines['bottom'].set_visible(False)
    ax.spines['left'].set_visible(False)
    # http://stackoverflow.com/a/11417222/190597 (gcalmettes)
    ax.xaxis.set_ticks_position('none')
    ax.yaxis.set_ticks_position('none')
    autolabel(patches, text_offset)

def autolabel(rects, shift=0.3):
    """
    http://matplotlib.org/1.2.1/examples/pylab_examples/barchart_demo.html
    """
    # attach some text labels
    for rect in rects:
        height = rect.get_height()
        if height > 0:
            plt.text(rect.get_x()+rect.get_width()/2., height+shift, '%d'%int(height),
                     ha='center', va='bottom')

x = [6,0,0,26,0,0,0,0,5,0,7,0,12,12,0,0,0,3,0,5,5,0,10,4,3,5,1,0,2,0,0,1,0,8,0,
     3,7,1,0,0,0,1,1,0,0,0,0,0,7,16,0,0,0,5,41]
fig, ax = plt.subplots(figsize=(14,5))
# make_hist(ax, x)
# make_hist(ax, [1,1,1,0,0,0], extra_y=1, text_offset=0.1)
make_hist(ax, x, bins=list(range(10))+list(range(10,41,5))+[np.inf], extra_y=6)
plt.show()

enter image description here

make_hist attempts to identify if all the values in x are integers. If so, it uses integer-based bin labels. For example, the bin label 10-14 represents the range [10, 14] (inclusive).

If, on the other hand, x contains floats, then make_hist will use half-open float-based bin labels. For example, 10-15 would represent the half-open range [10, 15).