[Tutor] sample dictionairies

[Steven D'Aprano]

On Sat, Apr 18, 2015 at 06:56:41PM -0700, Jim Mooney wrote:

> Where could I download Python sample dictionaries on different subjects.
> They're hard to type and I can only do small, limited ones to practice with.

I don't think you can download Python dicts. It would be a bit hard, 
since Python dicts exist only while the Python interpreter is running. 
Normally data is provided in text form, and then read from a file and 
converted into a dict or list as needed.

If you search the internet, you will find lots of places that you can 
download data on various topics. You can start at Wikidata:

https://www.wikidata.org/
‎http://en.wikipedia.org/wiki/Wikidata


But for simple experiments, there is an easier solution. Instead of 
typing the dicts yourself, create them programmatically!

Let's say I want a dict like:

{'a': 1, 'b': 2, 'c': 3, ... 'z': 26}

That would be a PITA to type out in full. But I can do this:

import string
pairs = zip(string.ascii_lowercase, range(1, 27))
d = dict(pairs)

zip is invaluable for this sort of thing, because it takes two (or more) 
independent sequences and "zips" them together into a single sequence of 
pairs. So zip('abcde...z', [1, 2, 3, 4, 5, ..., 26]) ends up as the 
sequence [('a', 1), ('b', 2), ('c', 3), ... ] which is exactly what 
dict() needs to build a dictionary.

Here's another example:

from random import random
d = dict((n, random()) for n in range(10000))

This uses a *generator expression* to provide a sequence of (key, value) 
pairs, where the keys are numbers 0, 1, 2, ... and the values are random 
floats. You might have seen list comprehensions:

[x**2+1 for x in range(15)]

for example. Generator expressions are similar, except they don't 
calculate the items up front into a list, they calculate them lazily 
only when needed. To turn a list comprehension into a generator 
expression, change the outer-most square brackets [ ] into round 
brackets:

(x**2+1 for x in range(15))

In Python 3, we also have *dict comprehensions* which are based on list 
comprehensions:

d = {n: random() for n in range(10000)}

So my advice is this: rather than look for somebody to provide you with 
ready-made dicts, learn how to assemble them yourself!

# dict of {name: age} pairs
names = 'Fred Wilma Barney Betty Homer Marge'.split()
ages = [35, 31, 34, 36, 30]
d = dict(zip(name, ages))


-- 
Steve