Mailing List Archive

On Mon, Aug 16, 1999 at 11:16:47PM -0600, Jerry Williams wrote:

> I have searched and searched the docs and am really frustrated with
> them.
> All I can seem to find is bits and pieces that hint about dictionaries.
> I finally found a table under mappings that listed the methods. But
> I couldn't find anything that would list each data type and what methods
> are available. What I am trying to do is take a file with a bunch of
> entries that appear more than once and list the count of each occurance
> in order of the highest count. So if the file looked like:
> a
> a
> a
> b
> b
> c
> Then is would print:
> 3 a
> 2 b
> 1 c

> I thought a dictionary would work great, but can't seem to figure out
> how to sort it by values and also get the key.
> list = dict.values()
> list.sort()

> Then I get lost. And the counts aren't going to be unique.
> Thanks for any help. Please point me to where the methods are listed
> for
> the dictionary. And some examples of dictionaries would be helpful.

Have you tried the Python Tutorial ?
http://www.python.org/doc/current/tut/

As for examples, here's probably what you want:

[thomas@tilburg ~]$ python
Python 1.5.1 (#1, Sep 3 1998, 22:51:17) [GCC 2.7.2.3] on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam

>>> dict = {} # an empty dictionary

>>> dict['spam'] = 'eggs' # Insert a string, with a string as key
>>> dict[5.4] = [2, 3, 4] # Insert a list, with a floating point number as key
>>> dict[1,2,3] = {} # insert an (empty) dictionary, with a tuple as key
>>> dict['me'] = dict # insert a reference to the dict itself

>>> dict # 'print' the dict
{'spam': 'eggs', (1, 2, 3): {}, 'me': {...}, 5.4: [2, 3, 4]}
>>> dict.keys() # print the dict's keys
['spam', (1, 2, 3), 'me', 5.4]
>>> dict.values() # print the dict's values.
[.'eggs', {}, {'spam': 'eggs', (1, 2, 3): {}, 'me': {...}, 5.4: [2, 3, 4]}, [2, 3, 4]]

>>> dict['spam'] # Index by string
'eggs'
>>> myindx = 5.4
>>> dict[myindx] # Index by variable
[2, 3, 4]
>>> dict["me"]["me"][1,2,3] # Index-index-index
{}

>>> for key in dict.keys(): # Print all dict values by hand
... print dict[key]
...
eggs
{}
{'spam': 'eggs', (1, 2, 3): {}, 'me': {...}, 5.4: [2, 3, 4]}
[2, 3, 4]

What you want to do is probably something like this:

import sys

if len(sys.argv) < 2:
print "Usage: %s <file>"%sys.argv[0]
sys.exit()

result = {}

file = open(sys.argv[1], 'r')
for line in file.readlines():
line = line[:-1] # We want to lose the newline
if result.has_key(line): # This could be done with exceptions too.
result[line] = result[line] + 1
else:
result[line] = 1

resultlines = result.keys()
resultlines.sort(lambda x, y: cmp(result[y], result[x]))
for line in resultlines:
print "%d %s"%(result[line], line)

Note cmp(result[y], result[x]), not cmp(result[x], result[y]), to sort the
list highest value first. You can also do a 'resultlines.reverse()' after
the sort to get the same effect.

This is what it does:

% python mysort.py testfile
66 d
60 f
36 l
36 ;
24 u
24 k
24 n
24 j
12 e
12 i
12 h
12 2
12 0
12 r
12 w
12 v
6 ,
4 s
3 p
2 a
1 m

Nevertheless, reading the tutorial and parts of the Library Reference really
are really good ideas.

and-i-didnt-even-show-the-other-10-good-ways-to-do-it'ly y'rs.

--
Thomas Wouters <thomas@xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!

Jerry Williams <jwilliam@xmission.com> wrote:
: I have searched and searched the docs and am really frustrated with
: them.
: All I can seem to find is bits and pieces that hint about dictionaries.
: I finally found a table under mappings that listed the methods. But
: I couldn't find anything that would list each data type and what methods
: are available. What I am trying to do is take a file with a bunch of
: entries that appear more than once and list the count of each occurance
: in order of the highest count. So if the file looked like:
: a
: a
: a
: b
: b
: c
: Then is would print:
: 3 a
: 2 b
: 1 c
: I thought a dictionary would work great, but can't seem to figure out
: how to sort it by values and also get the key.
: list = dict.values()
: list.sort()
: Then I get lost. And the counts aren't going to be unique.
: Thanks for any help. Please point me to where the methods are listed
: for
: the dictionary. And some examples of dictionaries would be helpful.

How about:

def repr_sorted_dict(d):
"""Return a string representation of a dictionary,
sorted by value first."""
def cmpitems_by_value(a, b):
"""sort first by second element in sequence."""
v = cmp(a[1], b[1])
# if equal, then sort by first element in sequence
if v == 0:
v = cmp(a[0], b[0])
return v
import string
# retrieve the (key, value) pairs
items = d.items()
items.sort(cmpitems_by_value)
# display the pairs
l = []
for k, v in items:
l.append("%s: %s" % (`k`, `v`))
# now combine them as a dictionary display
return '{' + string.joinfields(l, ', ') + '}'

Then you can:
print repr_sorted_dict(my_values)

And if you really wanted to get fancy, you could make this function the
__repr__ method of a subclass of UserDict.UserDict.

from UserDict import UserDict
class ReverseSortedDict(UserDict):
__repr__ = lambda self: repr_sorted_dict(self.data)

(Although I'd rewrite it to use the UserDict interface rather than
self.data itself, but that's just me. :)

Even if this is not exactly what you want, the "pairs = dict.items();
pairs.sort(...)" should get you where you are going. :)

This is documented at:
http://www.python.org/doc/current/lib/typesmapping.html
http://www.python.org/doc/current/lib/typesseq-mutable.html

-Arcege