Finding Top 20 Tags on StackOverflow main site

so-tags.py

import json
import pandas as pd
import numpy as np
import xmltodict
import matplotlib.pyplot as plt
f = open("Tags.xml")
all_data = f.read()
o = xmltodict.parse(all_data)
df = pd.DataFrame.from_dict(o['tags']['row'])
df[['counts']] = df[['@Count']].astype(int)

df2 = df.sort_values(by=['counts'], ascending=False).head(20)[['counts', '@TagName']]


"""
In [53]: df2
        counts       @TagName
2      1067078     javascript
11     1025688           java
6       918586             c#
4       885422            php
703     800779        android
422     712360         jquery
10      542985         python
1       511091           html
7       431790            c++
19333   414394            ios
14      380535          mysql
3       372444            css
15      319001            sql
56      282582        asp.net
3199    253474    objective-c
2327    235532  ruby-on-rails
0       227675           .net
86      210953         iphone
5       210835              c
64      170559         arrays
"""

counts = df2['counts'].as_matrix()
x = np.array(range(len(counts)))
labels = df2['@TagName'].values

fig = plt.figure()
fig.set_size_inches(15, 10.5)
ax = fig.add_axes([0.1, 0.1, 0.8, 0.8])
ax.set_xlabel("Tags")
ax.set_ylabel("Counts")
ax.bar(x, counts, align='center')
ax.set_xticks(x)
ax.set_xticklabels(labels)
fig.show()
fig.savefig('plot.png', format='png')