Creating dictionary dynamically in Python

Question

I have this list and number:

list = ['B','C']

The outcome that I need for my table is:

B    C    Prob
0    0    x
0    1    x
1    0    x
1    1    x

How can I build this truth table (there can be more vairables, not only 3) and assign a number to that row's probability?

I need to build it with a dictionary, I tried with some list comprehension but I don't know how to generate dynamically the truth table, considering that there can be more/less than 3 variables.

EDIT: to be more clear my goal is to have a dictionary like this:

dict = {"B":0/1,"C":0/1,"Prob":arbitraryNumber}

and I need to insert all these dictionaries into a list to represent the structure of a table, is it clearer now?

Thank you very much

Follow-up of http://stackoverflow.com/questions/23516212/remove-element-in-list-using-list-comprehension-python/23516225#23516225 :) — sshashank124, May 07 '14 at 11:35
@PadraicCunningham I have some data from a XML file, no calculation needed — The Condor, May 07 '14 at 12:47

Padraic Cunningham · Accepted Answer · 2014-05-07T16:14:39.673

1

You can generate the truth table using a powerset,

def power_set(items):
    n = len(items)
    for i in xrange(2**n):
        combo = []
        for j in xrange(n):
            if (i >> j) % 2 == 1:
                combo.append(1)
            else:
                combo.append(0)
        yield combo    # if you want tuples, change to yield tuple(combo)


In [13]: list(power_set(l))
Out[13]: [[0, 0], [1, 0], [0, 1], [1, 1]]

In [14]: l=['B','C','E']

In [15]: list(power_set(l))
Out[15]: 
[[0, 0, 0],
[1, 0, 0],
 [0, 1, 0],
 [1, 1, 0],
 [0, 0, 1],
 [1, 0, 1],
 [0, 1, 1],
 [1, 1, 1]]

If you want to make a dict of the data, change yield combo to yield tuple(combo)

Then you can store key value pairings like:

d={}
for data in power_set(l):
    d[data]="your_calc_prob"
print d
{(0, 1): 'your_calc_prob', (1, 0): 'your_calc_prob', (0, 0): 'your_calc_prob', (1, 1): 'your_calc_prob'}

If you want the output sorted you can use sorted() which makes a copy of the list and returns a list:

 sorted(list(power_set(l)))
 Out[21]: 
 [[0, 0, 0],
 [0, 0, 1],
 [0, 1, 0],
 [0, 1, 1],
 [1, 0, 0],
 [1, 0, 1],
 [1, 1, 0],
 [1, 1, 1]]

Or you can use the list method sort() which sorts the list in place:

In [22]: data = list(power_set(l))  
In [23]: data.sort()
In [24]: data
Out[24]: 
[[0, 0, 0],
[0, 0, 1],
[0, 1, 0],
[0, 1, 1],
[1, 0, 0],
[1, 0, 1],
[1, 1, 0],
[1, 1, 1]]

edited May 07 '14 at 16:14

answered May 07 '14 at 12:25

Padraic Cunningham

176,452
29
245
321

Cool! Perfect except for one thing: the numbers I get are "specular", I mean for instance that the second line should be 001 and not 100, can you explain me how to do it? – The Condor May 07 '14 at 13:51
Do you mean the output should be reversed? – Padraic Cunningham May 07 '14 at 14:43
I don't know the exact meaning of reverse in matrix speaking, but for instance I need [000,001,010,011,100,101,110,111], is it more clear now? – The Condor May 07 '14 at 14:50
Ah ok i understand. I will edit my answers when i get back on my computer – Padraic Cunningham May 07 '14 at 14:55
No problem.you should probably edit your question to reflect the output that you need. – Padraic Cunningham May 07 '14 at 15:05
let us [continue this discussion in chat](http://chat.stackoverflow.com/rooms/52239/discussion-between-the-condor-and-padraic-cunningham) – The Condor May 07 '14 at 20:29

msvalkon · Answer 2 · 2014-05-07T12:33:43.017

You can use itertools.product() to generate the truth table and then depending on the logical operation, determine the probability. I don't know which logical operation you would like to use so let's just create a dictionary each row:

>>> l = ['B', 'C']
>>> truth_table = [dict(zip(l, x)) for x in product((0, 1), repeat=2)]
>>> print(truth_table)
[{'B': 0, 'C': 0}, {'B': 0, 'C': 1}, {'B': 1, 'C': 0}, {'B': 1, 'C': 1}]

For calculating the probability, you'll probably need a separate function to do that. For example a logical disjunction for two keys with 0 and 1 being the values is basically equivalent to max().

>>> l.append('Prob')
>>> truth_table = [dict(zip(l, x + (max(x), )) for x in product((0, 1), repeat=2)]
>>> print(truth_table)
[{'B': 0, 'C': 0, 'Prob': 0},
 {'B': 0, 'C': 1, 'Prob': 1},
 {'B': 1, 'C': 0, 'Prob': 1},
 {'B': 1, 'C': 1, 'Prob': 1}]

Creating dictionary dynamically in Python

2 Answers2