“Got 1 columns instead of …” error in numpy

问题

I'm working on the following code for performing Random Forest Classification on train and test sets;

from sklearn.ensemble import RandomForestClassifier
from numpy import genfromtxt, savetxt

def main():
    dataset = genfromtxt(open('filepath','r'), delimiter=' ', dtype='f8')   
    target = [x[0] for x in dataset]
    train = [x[1:] for x in dataset]
    test = genfromtxt(open('filepath','r'), delimiter=' ', dtype='f8')

    rf = RandomForestClassifier(n_estimators=100)
    rf.fit(train, target)
    predicted_probs = [[index + 1, x[1]] for index, x in enumerate(rf.predict_proba(test))]

    savetxt('filepath', predicted_probs, delimiter=',', fmt='%d,%f', 
            header='Id,PredictedProbability', comments = '')

if __name__=="__main__":
    main()

However I get the following error on execution;

---->      dataset = genfromtxt(open('C:/Users/Saurabh/Desktop/pgm/Cora/a_train.csv','r'), delimiter='', dtype='f8')

ValueError: Some errors were detected !
    Line #88 (got 1435 columns instead of 1434)
    Line #93 (got 1435 columns instead of 1434)
    Line #164 (got 1435 columns instead of 1434)
    Line #169 (got 1435 columns instead of 1434)
    Line #524 (got 1435 columns instead of 1434)
...
...
...

Any suggestions as to how avoid it?? Thanks.

回答1:

genfromtxt will give this error if the number of columns is unequal.

I can think of 3 ways around it:

1. Use the usecols parameter

np.genfromtxt('yourfile.txt',delimiter=',',usecols=np.arange(0,1434))

However - this may mean that you lose some data (where rows are longer than 1434 columns) - whether or not that matters is down to you.

2. Adjust your input data file so that it has an equal number of columns.

3. Use something other than genfromtxt:

.............like this

回答2:

You have too many columns in one of your rows. For example

>>> import numpy as np
>>> from StringIO import StringIO
>>> s = """
... 1 2 3 4
... 1 2 3 4 5
... """
>>> np.genfromtxt(StringIO(s),delimiter=" ")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib64/python2.6/site-packages/numpy/lib/npyio.py", line 1654, in genfromtxt
    raise ValueError(errmsg)
ValueError: Some errors were detected !
    Line #2 (got 5 columns instead of 4)

回答3:

An exception is raised if an inconsistency is detected in the number of columns.A number of reasons and solutions are possible.

Add invalid_raise = False to skip the offending lines.

dataset = genfromtxt(open('data.csv','r'), delimiter='', invalid_raise = False)
If your data contains Names, make sure that the field name doesn’t contain any space or invalid character, or that it does not correspond to the name of a standard attribute (like size or shape), which would confuse the interpreter.

deletechars

Gives a string combining all the characters that must be deleted from the name. By default, invalid characters are ~!@#$%^&*()-=+~\|]}[{';: /?.>,<.

excludelist

Gives a list of the names to exclude, such as return, file, print… If one of the input name is part of this list, an underscore character ('_') will be appended to it.

case_sensitive

Whether the names should be case-sensitive (case_sensitive=True), converted to upper case (case_sensitive=False or case_sensitive='upper') or to lower case (case_sensitive='lower').

data = np.genfromtxt("data.txt", dtype=None, names=True,\
       deletechars="~!@#$%^&*()-=+~\|]}[{';: /?.>,<.", case_sensitive=True)

Reference: numpy.genfromtxt

回答4:

I had this error. The cause was a single entry in my data that had a space. This caused it to see it as an extra row. Make sure all spacing is consistent throughout all the data.

回答5:

It seems like the header that includes the column names have 1 more column than the data itself (1435 columns on header vs. 1434 on data).

You could either:

1) Eliminate 1 column from the header that doesn't make sense with data

2) Use the skip header from genfromtxt() for example, np.genfromtxt('myfile', skip_header=*how many lines to skip*, delimiter=' ') more information found in the documentation.

回答6:

In my case, the error aroused due to having a special symbol in the row.

Error cause: having special characters like

'#' hash
',' given the fact that your ( delimiter = ',' )

Example csv file

1,hello,#this,fails
1,hello,',this',fails

-----CODE-----

import numpy as numpy data = numpy.genfromtxt(file, delimiter=delimeter) #Error

Environment Note:

OS: Ubuntu

csv editor: LibreOffice

IDE: Pycharm

回答7:

I also had this error when I was also trying to load a text dataset with genfromtext and do text classification with Keras.

The data format was: [some_text]\t[class_label]. My understanding was that there are some characters in the 1st column that somehow confuse the parser and the two columns cannot be split properly.

data = np.genfromtxt(my_file.csv, delimiter='\t', usecols=(0,1), dtype=str);

this snippet created the same ValueError with yours and my first workaround was to read everything as one column:

data = np.genfromtxt(my_file, delimiter='\t', usecols=(0), dtype=str);

and split the data later by myself.

However, what finally worked properly was to explicitly define the comment parameter in genfromtxt.

data = np.genfromtxt(my_file, delimiter='\t', usecols=(0,1), dtype=str, comments=None);

According to the documentation:

The optional argument comments is used to define a character string that marks the beginning of a comment. By default, genfromtxt assumes comments='#'. The comment marker may occur anywhere on the line. Any character present after the comment marker(s) is simply ignored.

the default character that indicates a comment is '#', and thus if this character is included in your text column, everything is ignored after it. That is probably why the two columns cannot be recognized by genfromtext.

来源：https://stackoverflow.com/questions/23353585/got-1-columns-instead-of-error-in-numpy

标签

python

numpy

genfromtxt