Trying to append arrays

Question

I have this piece of code:

csvData = np.array([]);
csvData = np.append(csvData, ['1', '2016-01-01', 'Some text'])
csvData = np.append(csvData, ['2', '2016-01-02', 'Some more text'])
print(csvData)

it outputs: ['1' '2016-01-01' 'Some text' '2' '2016-01-02' 'Some more text']

I would like to get something like: [['1' '2016-01-01' 'Some text'], ['2' '2016-01-02' 'Some more text']]

I tried wrapping my data string into []. The reason I am doing this, because I want to collect each csv row, sort it by specific column and iterate data rows.

Any other solution is welcome.

I would sort using something like: csvData [np.argsort(csvData [:,2])]

Learn to use `np.conacatenate` before trying to use `np.append` as though it were a list append clone. You need to understand array dimensions when using `numpy`. — hpaulj, Nov 19 '16 at 15:49

score 3 · Accepted Answer · answered Nov 19 '16 at 14:22

Your desired output is a list of list which can be generated by the below code:

csvData = []
csvData.append(['1', '2016-01-01', 'Some text'])
csvData.append(['2', '2016-01-02', 'Some more text'])
#csvData would be
#[['1', '2016-01-01', 'Some text'], ['2', '2016-01-02', 'Some more text']]

But since you use numpy library, for converting it to the appropriate type this line helps:

csvData = np.asarray(csvData)
#csvData would be
[['1' '2016-01-01' 'Some text']
 ['2' '2016-01-02' 'Some more text']]

score 1 · Answer 2 · answered Nov 19 '16 at 14:19

1

You can try to save np.append in subarray and then append it to main array. Something like that:

array = []
subarray = ['1', '2016-01-01', 'Some text']
...
array.append(subarray)

Don't know what object np is.

answered Nov 19 '16 at 14:19

matox

885
11
27

1

`np` is the `numpy` package - numerical arrays; see the tag – hpaulj Nov 19 '16 at 15:45

score 1 · Answer 3 · edited May 23 '17 at 12:07

Before I give the solution please take look at why you should not build not repeatedly append to numpy arrays.

Having said that if you must do it (it very very inefficient and highly discouraged) here is how it can be done

>>> import numpy as np
>>> r1 = ['1', '2016-01-01', 'Some text']
>>> r2 = ['2', '2016-01-02', 'Some more text']
>>> ar = np.empty(shape=(0,3))
>>> ar = np.vstack( [ar, r1] )
>>> ar = np.vstack( [ar, r2] )
>>> ar
array([['1', '2016-01-01', 'Some text'],
       ['2', '2016-01-02', 'Some more text']], 
      dtype='<U32')

note: pands library built over numpy may be more suitable for your needs. It provides function to read CSV files pandas.read_csv

Trying to append arrays

3 Answers3