Reputation: 42329
I'm coming from this question Get information out of sub-lists in main list elegantly, where a great solution was proposed (two actually).
The issue I have right now is that the answer given assumes a certain number of elements in a list (I never said the number of elements could change so it was my fault really) to generate the final list.
This is the MWE
of the code:
a = [[0,1.1, 0.5, 99,99,0.7], [0,0.3, 1.4,99,99, 0.2], [1,0.6, 0.2,99,99, 1.], [1,1.1, 0.5,99,99, 0.3], [2,0.2, 1.1,99,99, 0.8], [2,1.1, 0.5,99,99, 1.], [3,1.2, 0.3,99,99, 0.6], [3,0.6, 0.4,99,99, 0.9], [4,0.6, 0.2,99,99, 0.5], [4,0.7, 0.2,99,99, 0.5]]
a_processed = []
from collections import defaultdict
nums = defaultdict(list)
for arr in a:
key = tuple(arr[1:5]) # make the first two floats the key
nums[key].append( arr[5] ) # append the third float for the given key
a_processed = [[k[0], k[1], k[2], k[3], round(sum(vals)/len(vals),2)] for k, vals in nums.items()]
which returns:
a_processed = [[0.7, 0.2, 99, 99, 0.5], [0.3, 1.4, 99, 99, 0.2], [0.6, 0.2, 99, 99, 0.75], [0.6, 0.4, 99, 99, 0.9], [1.2, 0.3, 99, 99, 0.6], [0.2, 1.1, 99, 99, 0.8], [1.1, 0.5, 99, 99, 0.67]]
I want to replace the last line which explicitly names the elements k[0], k[1], k[2], k[3]
to something more general. I tried simply using:
a_processed = [[k, round(sum(vals)/len(vals),2)] for k, vals in nums.items()]
but that stores k
as a tuple inside each sub-list:
a_processed = [[(0.7, 0.2, 99, 99), 0.5], [(0.3, 1.4, 99, 99), 0.2], [(0.6, 0.2, 99, 99), 0.75], [(0.6, 0.4, 99, 99), 0.9], [(1.2, 0.3, 99, 99), 0.6], [(0.2, 1.1, 99, 99), 0.8], [(1.1, 0.5, 99, 99), 0.67]]
So how can I generalize that line to any number of elements without having to explicitly write down each one?
Upvotes: 1
Views: 325
Reputation: 13372
If I got your question right, the following is what you seek.
a_processed = [list(k)[:4] + [round(sum(vals)/len(vals),2)] for k, vals in nums.items()]
The + operator joins the 2 lists in given order.
Upvotes: 2
Reputation: 5732
Convert the tuple to a list using list(k)
, then extend that list with whatever other elements you need it to contain.
(You can concatenate lists in python using the +
operator.)
a_processed = [list(k) + [round(sum(vals)/len(vals),2)] for k, vals in nums.items()]
Upvotes: 6