Reading row elements in a new array in python

Question

I am converting code to write a function from a different data type.

The original code was:

        note_inf_track = np.array([(n.note, n.onset/div, n.duration/div, n.velocity, n.channel, track_nr)
                                    for n in m.tracks[track_nr].notes], 
                                  dtype = [('pitch', np.int), 
                                           ('onset', np.float), 
                                           ('duration', np.float), 
                                           ('velocity', np.int),
                                           ('channel', np.int),
                                           ('track', np.int)])

Now my input data is a 2-dimensional list, I am not working with notes anymore.

 for line in lines:
      #print("b");
      element = [];
      for x in line.split(','):
         element.append(x.strip('\r\n'));
      elements.append(element);


  note_inf_track = np.array([(round((round(np.asarray(elements[2], dtype="float")))), (round(np.asarray(elements[0], dtype="float"))),(round(np.asarray(elements[:][1], dtype="float"))))], 
                  dtype = [('pitch', np.int), 
                    ('onset', np.float), 
                    ('duration', np.float)])

I am struggling to add the columns at once.

elements[2] seems to give me the row instead of the column. I can't seem to replace the for loop. Maybe my syntax is all off, I am used to java and c++, fairly new to Python.

--Update--

Based on Tarun Gaba's answer, I tried this:

    note_inf_track = np.array([((round(el[2])), float(el[0]),float(el[1])) for el in elements],
                dtype = [('pitch', np.int)
                ('onset', np.float),
                ('duration', np.float)]);

Gives me an error:

note_inf_track = np.array([((round(el[2])), float(el[0]),float(el[1])) for el in elements],
TypeError: a float is required

Here is the output of print(elements):

[['0.066667', ' 0.200000', ' 50.180000', ' 0.000644'], ['0.266667', ' 0.266667', ' 59.180000', ' 0.006583'], ['0.550000', ' 0.366667', ' 59.180000', ' 0.002129'], ['0.933333', ' 0.350000', ' 59.180000', ' 0.005972'], ['1.316667', ' 0.050000', ' 59.180000', ' 0.010053'], ['1.366667', ' 0.166667', ' 61.180000', ' 0.008109'], ['1.550000', ' 0.233333', ' 61.180000', ' 0.009170'], ['1.783333', ' 0.416667', ' 63.180000', ' 0.023811'], ['2.250000', ' 0.166667', ' 63.180000', ' 0.016253'], ['2.416667', ' 0.850000', ' 64.180000', ' 0.019314'], ['3.300000', ' 0.116667', ' 64.180000', ' 0.018684'], ['3.433333', ' 0.133333', ' 64.180000', ' 0.016786'], ['3.583333', ' 0.333333', ' 63.180000', ' 0.008623'], ['4.816667', ' 0.383333', ' 63.180000', ' 0.036858'], ['5.200000', ' 0.166667', ' 61.180000', ' 0.006060'], ['5.366667', ' 0.366667', ' 63.180000', ' 0.010417'], ['5.783333', ' 0.333333', ' 63.180000', ' 0.008371'], ['6.116667', ' 0.383333', ' 64.180000', ' 0.007488'], ['6.533333', ' 0.233333', ' 64.180000', ' 0.014582'], ['6.766667', ' 0.333333', ' 63.180000', ' 0.004457'], ['7.533333', ' 0.516667', ' 61.180000', ' 0.004700'], ['8.050000', ' 0.316667', ' 63.180000', ' 0.006959'], ['8.366667', ' 0.300000', ' 64.180000', ' 0.013522'], ['8.666667', ' 0.166667', ' 63.180000', ' 0.008083'], ['8.833333', ' 0.150000', ' 64.180000', ' 0.010620'], ['8.983333', ' 0.250000', ' 63.180000', ' 0.004493'], ['9.233333', ' 0.116667', ' 64.180000', ' 0.012834'], ['9.350000', ' 0.333333', ' 63.180000', ' 0.005321'], ['9.716667', ' 0.300000', ' 64.180000', ' 0.006902'], ['10.033333', ' 0.183333', ' 63.180000', ' 0.002515'], ['10.216667', ' 0.133333', ' 62.180000', ' 0.005928'], ['10.350000', ' 0.600000', ' 63.180000', ' 0.004920'], ['10.950000', ' 0.133333', ' 64.180000', ' 0.006754'], ['11.083333', ' 0.116667', ' 63.180000', ' 0.003831'], ['11.200000', ' 0.316667', ' 62.180000', ' 0.002493']]

You could just transform elements to numpy array elements = np.array(elements) and then elements[:, 2] will select the 3nd column for you. — Imanol Luengo
– Imanol Luengo, Commented Dec 3, 2014 at 11:56
If I do that, then it gives me: TypeError: only length-1 arrays can be converted to Python scalars — dorien
– dorien, Commented Dec 3, 2014 at 12:00

Tarun Gaba · Accepted Answer · 2014-12-03 16:54:06Z

2

elements is a list of lists here. To access 3rd column(as what you seem to be trying by elements[2]), you need to do something like this:

elements = [[1,2,3], \
            [4,5,6], \
            [7, 8, 9]]

column  = [i[2] for i in elements]
print column
#[3,6,9]

For your case, It should be something on the lines of:

np.array([el[2] for el in elements], [float(el[0]) for el in elements], [float(el[1])) for el in elements], dtype= .....

edited Dec 3, 2014 at 16:54

answered Dec 3, 2014 at 11:52

Tarun Gaba

1,1131 gold badge9 silver badges16 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

dorien Over a year ago

Based on this, I have changed my code, see updated question. I get an error that a float is required now. Getting closer perhaps...

Tarun Gaba Over a year ago

I think, you are using round for non-float. A round is not defined for non-floats in python. try removing the round() from round(el[2])

dorien Over a year ago

When I remove the round(), I get: TypeError: 'tuple' object is not callable

Tarun Gaba Over a year ago

You should be creating three lists independently. I have updated my answer. Hope it helps

dorien Over a year ago

Adding this fixed it element.append(float(x.strip('\r\n')));

Imanol Luengo · Accepted Answer · 2014-12-03 18:26:42Z

1

The problem is that your data is read as list of strings.

Modify your code from:

element.append(x.strip('\r\n'));

To:

element.append(float(x.strip('\r\n')));

To have your data as floats. You could also use round(float(...)) if you need rounded data.

Then put the data into a numpy array:

>>> import numpy as np
>>> data = np.array(elements)

And access to the columns as data[:, column_idx], e.g. for column 3:

>>> data[:, 2]

edited Dec 3, 2014 at 18:26

answered Dec 3, 2014 at 18:20

Imanol Luengo

16k3 gold badges52 silver badges68 bronze badges

Collectives™ on Stack Overflow

Reading row elements in a new array in python

2 Answers 2

5 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

5 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related