How to change multiple values in a list of lists from str to float-CodePudding

I have the following list:

[[
     ['3.31497114423,50.803721015'],
     ['7.09205325687,50.803721015'],
     ['7.09205325687,53.5104033474'],
     ['3.31497114423,53.5104033474'], 
     ['3.31497114423,50.803721015']
         ]]

How can I convert the list so that each of the value is a floating point?

Wanted result:

[[
     [3.31497114423,50.803721015],
     [7.09205325687,50.803721015],
     [7.09205325687,53.5104033474],
     [3.31497114423,53.5104033474],
     [3.31497114423,50.803721015]
         ]]

CodePudding user response：

You can use list comprehension and split string base ',' then use map(float, list) for converting each str in list to float.

res = [[list(map(float,l[0].split(','))) for l in lst] for lst in lst_lst]

print(res)

[[
    [3.31497114423, 50.803721015], 
    [7.09205325687, 50.803721015], 
    [7.09205325687, 53.5104033474], 
    [3.31497114423, 53.5104033474], 
    [3.31497114423, 50.803721015]
]]

Explanation:

>>> '3.31497114423,50.803721015'.split(',')
['3.31497114423', '50.803721015']

>>> list(map(float, '3.31497114423,50.803721015'.split(',')))
[3.31497114423, 50.803721015]

CodePudding user response：

You can just eval the string:

a=[[
     ['3.31497114423,50.803721015'],
     ['7.09205325687,50.803721015'],
     ['7.09205325687,53.5104033474'],
     ['3.31497114423,53.5104033474'], 
     ['3.31497114423,50.803721015']
         ]]
a=[[list(eval(i[0])) for i in a[0]]]
print(a)

Output:

[[[3.31497114423, 50.803721015], [7.09205325687, 50.803721015], [7.09205325687, 53.5104033474], [3.31497114423, 53.5104033474], [3.31497114423, 50.803721015]]]

CodePudding user response：

x = [[['3.31497114423,50.803721015'],
['7.09205325687,50.803721015'],
['7.09205325687,53.5104033474'],
['3.31497114423,53.5104033474'],
['3.31497114423,50.803721015']]]
    
y = [[]]
for item in x[0]:
    t = []
    z = item[0].split(',')
    for f in z:
        t.append(float(f))
    y[0].append(t)
        
print(y)

CodePudding user response：

You could also use this monstrosity which uses only one list comprehension:

[list(map(float,p[0][i][0].split(","))) for i in range(len(p[0]))]

, where p is your original array. The solution of Ash Nazg is more elegant though. But be careful with using eval() on untrusted data.

CodePudding user response：

I'd suggest breaking the task into two simpler ones. First you want some code that turns a string containing comma-separated floats into a list of floats, something like:

def parse_csv_floats(s):
    return list(map(float, s.split(",")))

parse_csv_floats("3.1,50.8")  # => [3.1, 50.8]

and then you want to call this code for each element of your egregiously nested structure:

deep_csv_floats = [[
     ['3.31497114423,50.803721015'],
     ['7.09205325687,50.803721015'],
     ['7.09205325687,53.5104033474'],
     ['3.31497114423,53.5104033474'], 
     ['3.31497114423,50.803721015']
]]
deep_floats = [
    [parse_csv_floats(csv) for [csv] in inner]
    for inner in deep_csv_floats
]

which stores your expected output in deep_floats. Note that the for [csv] in inner will ensure that the CSVs are always exactly one string in a list. If you want to parse values that are more nested, you will want another nested list comprehension.