CODEWITHSUNDEEP
pythonpython-3.xlistnumpystatistics
vahid sabet
vahid sabet

Reputation: 583

Python calculate percentage of list values one by others

I have a list of values and I want to get difference percentage of the each element with others.

  x[1] with x[2],x[3],x[4]
  x[2] with x[1],x[3],x[4] and so on.
    
x =["3150000", "3156000", "3150000","3250000"]

Here is what I want using python:

((3150000 - 3156000) / 3150000) * 100

Desired output:

[-0.19, 0, 3.17, 0.19, 0.19 ..., 3.08]

What's the most efficient way to do this?

Upvotes: 0

Views: 595

Answers (3)

Andreas
Andreas

Reputation: 9197

You can use itertools.permutations:

import itertools

def func(l):
    return round(((l[0] - l[1]) / l[0]) * 100)

lst = ["3150000", "3156000", "3150000","3250000"]

# map the string values in list to their integer counterpart
lst = map(int, lst)

# create permutations of length 2 from list
permutations = list(itertools.permutations(lst, 2))

# apply function to the permutations
results = {l: func(l) for l in permutations}

# print results
print(*results.items(), sep='\n')

((3150000, 3156000), -0.19)
((3150000, 3150000), 0.0)
((3150000, 3250000), -3.17)
((3156000, 3150000), 0.19)
((3156000, 3250000), -2.98)
((3250000, 3150000), 3.08)
((3250000, 3156000), 2.89)

As one-liner:

results = {l: func(l) for l in list(itertools.permutations(map(int, lst), 2))}

Upvotes: 0

Ivan
Ivan

Reputation: 40668

Assuming you are looking for all possible difference pairs.

You can compute the matrix of differences with broadcasting, then divide back and normalize to a percentage:

>>> (x - x[:, None])/x*100
array([[ 0.        ,  0.19011407,  0.        ,  3.07692308],
       [-0.19047619,  0.        , -0.19047619,  2.89230769],
       [ 0.        ,  0.19011407,  0.        ,  3.07692308],
       [-3.17460317, -2.97845374, -3.17460317,  0.        ]])

Where component at position i, j corresponds to (x[i] - x[j]) / x[i] * 100.

I did preprocess x with:

>>> x = np.array([int(i) for i in x])
array([3150000, 3156000, 3150000, 3250000])

Similarly you could use np.array(x).astype(float).

Upvotes: 0

alexzander
alexzander

Reputation: 1875

is this what you want?

x =["3150000", "3156000", "3150000","3250000"]
x = list(map(int, x))

for x1 in x:
    for x2 in x:
        if x1 == x2:
            continue
                
        diff_per = abs(x1 - x2) / x1 * 100
        print("abs({x1} - {x2}) / {x1} * 100 == {:.2f}%".format(diff_per, x1=x1, x2=x2))

output:

abs(3150000 - 3156000) / 3150000 * 100 == 0.19%
abs(3150000 - 3250000) / 3150000 * 100 == 3.17%
abs(3156000 - 3150000) / 3156000 * 100 == 0.19%
abs(3156000 - 3150000) / 3156000 * 100 == 0.19%
abs(3156000 - 3250000) / 3156000 * 100 == 2.98%
abs(3150000 - 3156000) / 3150000 * 100 == 0.19%
abs(3150000 - 3250000) / 3150000 * 100 == 3.17%
abs(3250000 - 3150000) / 3250000 * 100 == 3.08%
abs(3250000 - 3156000) / 3250000 * 100 == 2.89%
abs(3250000 - 3150000) / 3250000 * 100 == 3.08%

Upvotes: 1

Related Questions