CODEWITHSUNDEEP
pythonlistpython-3.xloopsiteration
xyzo
xyzo

Reputation: 31

Combining two lists and printing all the values corresponding to a certain value

I need to combine two lists and then count all the values corresponding to a certain value.

The two lists are:

inControl = ["False", "False", "True", "True","False", "True", "False", "True", "True", "False", "False", "False", "False", "False", "False", "True", "False", "True", "False", "False"]

rts = [379, 396, 480, 443, 365, 280, 487, 446, 350, 367, 405, 391, 484, 359, 367, 305, 359, 479, 436, 333]

I need to sum all of the rts corresponding to all of the False values, and then the same for True values (they are all in order).

I've basically got as far as combining the two lists using as a zip function, but am completely lost as to what to do next... any help would be appreciated.

Many thanks:)

Upvotes: 0

Views: 69

Answers (4)

dawg
dawg

Reputation: 104092

You can produce tuples that have a single number either in the False position or True position in a list comprehension or generator exrpession:

>>> [(e,0) if c=='False' else (0,e) for e, c in zip(rts, inControl)]
[(379, 0), (396, 0), (0, 480), (0, 443), (365, 0), (0, 280), (487, 0), (0, 446), (0, 350), (367, 0), (405, 0), (391, 0), (484, 0), (359, 0), (367, 0), (0, 305), (359, 0), (0, 479), (436, 0), (333, 0)]

You can then sum series of tuples with reduce:

>>> reduce(lambda x, y: (x[0]+y[0], x[1]+y[1]), (((e,0) if c=='False' else (0,e) for e, c in zip(rts, inControl))))
(5128, 2783)

You can actually use False and True booleans to access the tuple:

>>> t=reduce(lambda x, y: (x[0]+y[0], x[1]+y[1]), ((e if c=='False' else 0, e if c=='True' else 0) for e, c in zip(rts, inControl)))
>>> t[False]
5128
>>> t[True]
2783

Or, you can use map if that makes more sense to you:

>>> map(sum, zip(*((e,0) if c=='False' else (0,e) for e, c in zip(rts, inControl))))
[5128, 2783]

Or, you can create a dict with the sums:

>>> dict(zip([False, True], map(sum, zip(*[(e,0) if c=='False' else (0,e) for e, c in zip(rts, inControl)]))))
{False: 5128, True: 2783}

If you have Pandas, a great way to do this is with .groupby() and sum:

>>> import pandas as pd
>>> df=pd.DataFrame({'rts':rts, 'inControl':inControl})
>>> df
   inControl  rts
0      False  379
1      False  396
2       True  480
3       True  443
4      False  365
5       True  280
6      False  487
7       True  446
8       True  350
9      False  367
10     False  405
11     False  391
12     False  484
13     False  359
14     False  367
15      True  305
16     False  359
17      True  479
18     False  436
19     False  333
>>> df.groupby(inControl).sum()
        rts
False  5128
True   2783

Upvotes: 0

Hai Vu
Hai Vu

Reputation: 40773

Given:

inControl = ["False", "False", "True", "True","False", "True", "False", "True", "True", "False", "False", "False", "False", "False", "False", "True", "False", "True", "False", "False"]
rts = [379, 396, 480, 443, 365, 280, 487, 446, 350, 367, 405, 391, 484, 359, 367, 305, 359, 479, 436, 333]

I assume that inControl is a list of strings, not a list of booleans. For my solution to work, I am going to convert inControl to list of booleans:

inControl = [element == 'True' for element in inControl]  # ==> [False, False, ...]

Use itertools.compress to calculate sum of all True elements:

import itertools
true_sum = sum(itertools.compress(rts, inControl))  # 2783

Now, we can calculate the false_sum:

grand_sum = sum(rts)              # 7911
false_sum = grand_sum - true_sum  # 5128

Upvotes: 0

jme
jme

Reputation: 20765

Jean-François's solution will work just fine and is pretty readable. It does make two passes over the data, though. If the list is small this isn't a big deal, but if it's big you can roughly halve the time of operation by taking a single pass.

One general approach is this:

totals = {}
for flag, value in zip(inControl, rts):
    totals[flag] = totals.setdefault(flag, 0) + value

This code does not assume that inControl has only False and True. It can in fact have any number of unique values.

A cuter way is to use the Counter class from the collections module. A Counter is a dictionary intended to keep track of counts. Adding two counters does the obvious thing: the values of identical keys are summed. We can create a Counter instance for each pair of elements and add up all of the counters. Note that creating a Counter for each element is probably overkill -- the above solution is more efficient. But for educational purposes, this solution looks like:

from collections import Counter
counters = (Counter({k: v}) for k, v in zip(inControl, rts))
sum(counters, Counter())

Upvotes: 2

Jean-François Fabre
Jean-François Fabre

Reputation: 140286

Seems weird that the booleans are strings, but...

Use zip, then sum on matching elements.

(re-create the zip for the other part in Python 3 because zip is an iterable)

inControl = ["False", "False", "True", "True","False", "True", "False", "True", "True", "False", "False", "False", "False", "False", "False", "True", "False", "True", "False", "False"]

rts = [379, 396, 480, 443, 365, 280, 487, 446, 350, 367, 405, 391, 484, 359, 367, 305, 359, 479, 436, 333]

z=zip(rts,inControl)
sf=sum(x[0] for x in z if x[1]=='False')
z=zip(rts,inControl)
st=sum(x[0] for x in z if x[1]=='True')
print(sf,st)

result:

5128 2783

Maybe st could be computed with less string comparisons with sum: st=sum(rts)-sf (more additions, less string comparisons)

variant: small loop for True & False

s=dict()

for c in ['False','True']:
    z=zip(rts,inControl)
    s[c]=sum(x[0] for x in z if x[1]==c)

Upvotes: 2

Related Questions