I have a .dat file which contains about 1000 lines each line contains
letter int int int int boolean
and i am trying to read it in so each line is a new row on my array.
currently I have np.genfromtxt('myfile.dat') which gives me
nan 23. 34. 23. 55. 1.
this is almost right but that nan should be the letter 't' any idea how I get to read in the correct letter? And also how do I get rid of the . after each number?
cheers
One way is defining a new dtype. For example:
import numpy as np
desc = np.dtype([('letter', 'S1'), ('v1', float), ('v2', float),
('v3', float), ('v4', float)])
and use in genfromtext:
data = np.genfromtxt(fobj, dtype=desc)
This file content:
x 23. 34. 23. 55. 1.
y 23. 34. 23. 55. 1.
would give you this data:
array([(b'x', 23.0, 34.0, 23.0, 55.0), (b'y', 23.0, 34.0, 23.0, 55.0)],
dtype=[('letter', 'S1'), ('v1', '<f8'), ('v2', '<f8'), ('v3', '<f8'), ('v4', '<f8')])
This is an record array. You can access one line:
>>> data[0]
(b'x', 23.0, 34.0, 23.0, 55.0)
or one column:
>>> data['letter']
array([b'x', b'y'],
dtype='|S1')
or one entry:
>>> data[0][1]
23.0
>>> data['v1'][1]
23.0
dtype=None. Either way the result will be a 1d structured array, one record per line.