Access element of list of arrays

user_12981887 0
Viewed 113

I have a list of data made up of arrays. I want to able to access each element of these arrays. However, when I try to access each element of one of the lists I instead get a whole list returned i.e. 

>>> vec_data[0]
'[array([129.1203,  65.5152, -15.7962]), array([ -75.8793,  -27.2767, -136.3468]), array([  98.54  , -110.7603,  187.9171]), array([ 83.9628, 126.839 , -61.4756]), array([ -63.1397, -216.348 , -165.1392]), array([ 32.8174,  45.1573, 146.6547]), array([65.2439, 82.8289, 13.3427]), array([  -8.481 ,  -38.3387, -138.1786]), array([-223.8824,   86.1375,   -2.5446]), array([211.8308,  51.894 ,  89.5078]), array([  21.4793, -177.2219,   -0.573 ])]\n'

I then tried vec_data[0][0] hoping to get [129.1203, 65.5152, -15.7962] returned but instead got: 

>>> vec_data[0][0]
'['

Can anyone provide any insight on how I may get back the output I want from this list?

I have attached a small excerpt of the data below to aid the problem. 

['[array([129.1203,  65.5152, -15.7962]), array([ 83.094 , 135.5689,  57.7153]), array([  97.337 , -158.2453,   26.0811]), array([  31.3163,  121.7492, -100.4891]), array([ 60.797 ,  20.4243, -45.3821]), array([  -1.9805, -145.0236,  -57.6177]), array([116.4612,  18.3822, -99.8011]), array([38.402 , 13.7894, 85.094 ]), array([-321.057 ,  147.0128,  173.035 ]), array([  99.6488,  -74.9626, -110.7028]), array([ 134.3792, -110.4547,   -8.8729])]\n', '[array([ -67.876  ,  -28.46764, -118.16426]), array([ 58.368  , 107.64964,  90.12756]), array([ 102.6007, -165.1867,   26.1187]), array([  39.58716,  124.4238 , -106.68995]), array([ 58.19804,  18.7423 , -40.15905]), array([  -4.9479, -151.5177,  -56.3511]), array([ 99.946 ,  27.8355, -89.1489]), array([51.115, -3.302, 80.843]), array([-326.252 ,  170.7748,  168.0119]), array([ 120.9826,  -91.66  , -116.6699]), array([114.7974, -82.5404,   2.0438])]\n', '[array([ -28.196  ,  -14.69983, -149.5836 ]), array([ 54.399  , 108.75513, 106.2178 ]), array([  93.0723, -146.1217,   27.1268]), array([  15.1398,  104.9325, -128.3365]), array([ 89.2691,   2.3471, -21.0604]), array([ -15.2423, -155.4835,  -28.2395]), array([ 86.8051, -18.2533, -74.2126]), array([73.194 , 60.0037, 67.0507]), array([-314.68  ,  170.6532,  145.9454]), array([  69.4455,  -98.9357, -135.0111]), array([147.9465, -88.8198,  36.5976]

EDIT:

using the method suggested by revliscano: 

import re
import json


string = vec_data[0]
r = re.sub('array\((\[.+?\])\)', lambda x: x.groups()[0], string)
r = r.replace('\n', '')
list_ = json.loads(r)

Give me the desired output for vec_data[0] however, I want to get this to loop over all of vec_data - any suggestions?

python-3.x python
4 Answers

If you have a way of not putting this data into your program as a string, that would be ideal, but I believe that the following should work to convert your data into a better format.

import numpy as np
array = np.array
vec_data = eval(vec_data)

After doing this with your sample code I got a list of arrays.

If you want this to be in a list format instead, you could replace array = np.array with array = list, but by the looks of your input data, it was probably originally a numpy array.

user_133578780

Disclaimer: The following answer does not have anything to do with numpy arrays, but it finds a way to parse the string you have in vec_data[0], so that it can be accesible and used as a list, which at the end I think is what you are trying to do.

We're going to be using regex (through the re module) and the json module to achieve this.

If all the items of vec_data have similar values (strings that need to be parsed, just like vec_data[0]), then you can use this solution as a function, and use a list comprehension to iterate vec_data, calling that function on each item of it:

import re
import json


def parse_array(string):
    r = re.sub('array\((\[.+?\])\)', lambda x: x.groups()[0], string)
    r = r.replace('\n', '')
    list_ = json.loads(r)
    return list_


new_vec_data = [parse_array(string) for string in vec_data]

output for print(new_vec_data[0][0]):

[129.1203, 65.5152, -15.7962]
user_73034340

Try vec_data[0][0][0] to access the first element of first list

user_135950810

How about this?

import re
import json

output = []
for i in vec_data:
    string = vec_data[0]
    r = re.sub('array\((\[.+?\])\)', lambda x: x.groups()[0], string)
    r = r.replace('\n', '')
    output.append(json.loads(r))

print(output)

This way, you will get a list of data list.

If you want a single list with all the data, you can use extend instead such as,

import re
import json

output = []
for i in vec_data:
    string = vec_data[0]
    r = re.sub('array\((\[.+?\])\)', lambda x: x.groups()[0], string)
    r = r.replace('\n', '')
    output.extend(json.loads(r))

print(output)
user_43506410
Related