Get numbers from string python

I am facing the following problem. I am trying to download this dataset: Dataset link

in this way:

data_file_url = ''
D = np.array(pd.read_csv(data_file_url,header=0))
D = D[ np.random.choice(np.arange(D.shape[0]), D.shape[0], replace=False) ,:]
Dx = D[:,0:2]
Dy = D[:,2]

but it seems that is comes in a .txt array format. Thats not really the problem, but the string itself is. It comes in this form:

[['    665845    557965']
 ['    597173    575538']
 ['    618600    551446']
 ['    650661    861267']
 ['    599647    858702']
 ['    684091    842566']]

, where all the arrays are a giant wierd string with a lot of blanc spaces and two number which are the coordinates. I am trying to get it in this form [123124412, 12312442]

The dataset can be downloaded in .txt or .ts format.

I tried to split, then cast to int but I am getting all the number instead of 2, obviously.

Thanks for the help or advice!


Have you tried using the optional arguments of pd.read_csv?
Try the following:
D = np.array(pd.read_csv(data_file_url,header=0,delimiter=' '))