Initialize the Counter once, let the keys be artists, and augment a key (artist) each time through the loop:
c = Counter()
for d in entries:
arts = d['artist']
c[arts] += 1
print(c.most_common(10))
When arts
is a string, then c = Counter(arts)
counts the characters in arts
:
In [522]: collections.Counter('Led Zepplin')
Out[522]: Counter({'e': 2, 'p': 2, ' ': 1, 'd': 1, 'i': 1, 'L': 1, 'l': 1, 'n': 1, 'Z': 1})
In contrast:
In [523]: c = collections.Counter()
In [524]: c['Led Zepplin'] += 1
In [525]: c['The Rolling Stones'] += 1
In [526]: c.most_common()
Out[526]: [('Led Zepplin', 1), ('The Rolling Stones', 1)]
Alternatively, as Jon Clements points out, build a list of all the artists, and then count the list:
c = Counter(d['artist'] for d in entries)
print(c.most_common(10))
Note that the above uses a generator expression to avoid building a (possibly) large temporary list, and at the same time has a much more succinct, readable syntax.