[FIXED] Plotting frequency of elements within a list of lists

Issue

I have a list of lists:

my_list= [['UV'],
     ['SB'],
     ['NMR'],
     ['ISSN'],
     ['UK', 'USA'],
     ['MT'],
     ['UK'],
     ['UK'],
     ['ESP'],
     ['UK'],
     ['UK'],
     ['UK'],
     ['UK'],
     ['UK'],
     ['UK']]

that I would like to plot in terms of frequency (from the most frequent term to the less frequent).

I am finding some issue in counting the items. What I first did is to flatten the list of lists:

flattened = [] 
for sublist in my_list: 
    for val in sublist: 
        flattened.append(val) 

Then I tried to count items it

from collections import Counter
import pandas as pd

counts = Counter(flattened)
df_ver = pd.DataFrame.from_dict(counts, orient='index')
df_ver.plot(kind='bar')

However it does not work. Also it should be not sorted, I guess.

Solution

Let's try with pure Python:

counts = {}

for countries in my_list:
    for country in countries:
        counts[country] = counts.get(country,0) +1

sorted_counts = sorted(counts.items(), key=lambda i: (-i[1],i[0])) # sort by count and alphabetically if draw
# ktop = 10
# sorted_counts = sorted_counts[:ktop]
countries, counts = list(zip(*sorted_counts))

plt.bar(countries, counts);

Answered By - Sergey Bushmanov