ValueError: invalid literal for float():

[TMS]

I'm working on a linear regression program and when I try to convert the input file data to a float (it reads in as a string), I get this error. Perhaps someone can help me? I'm using Windows XP, and IDLE. (also the console screen).

#! /usr/bin/env python
import sys
import os
from numpy import * 
numLst = [] #an empty list
if len(sys.argv) == 2:
    infile = open(sys.argv[1], 'r')
elif len(sys.argv) <2:
    infile = raw_input("Name of file: ")
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
data = open(infile, 'r')
for x in data:
    x = float(x) # HERE IS THE PROBLEM
    numLst.append(x.strip('\n').split(" "))
 
a = array(numLst)
print a
 

Input file looks like this:

1.0 2.0
2.0 3.5
2.5 5.0
3.5 6.5

which are data points, x and y respectively.

Thanks!

TMS

[bvdet]

I'm working on a linear regression program and when I try to convert the input file data to a float (it reads in as a string), I get this error. Perhaps someone can help me? I'm using Windows XP, and IDLE. (also the console screen).

#! /usr/bin/env python
import sys
import os
from numpy import * 
numLst = [] #an empty list
if len(sys.argv) == 2:
    infile = open(sys.argv[1], 'r')
elif len(sys.argv) <2:
    infile = raw_input("Name of file: ")
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
data = open(infile, 'r')
for x in data:
    x = float(x) # HERE IS THE PROBLEM
    numLst.append(x.strip('\n').split(" "))
 
a = array(numLst)
print a
 

Input file looks like this:

1.0 2.0
2.0 3.5
2.5 5.0
3.5 6.5

which are data points, x and y respectively.

Thanks!

TMS

Each iteration reads a line from the file. The argument to float must be in the correct format. See below:

>>> x = '1.0 2.0'
>>> float(x)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
ValueError: invalid literal for float(): 1.0 2.0
>>> x,y = [float(i) for i in x.split()]
>>> x
1.0
>>> y
2.0
>>> 

Separate the numbers and pass each one to float().

[ilikepython]

I'm working on a linear regression program and when I try to convert the input file data to a float (it reads in as a string), I get this error. Perhaps someone can help me? I'm using Windows XP, and IDLE. (also the console screen).

#! /usr/bin/env python
import sys
import os
from numpy import * 
numLst = [] #an empty list
if len(sys.argv) == 2:
    infile = open(sys.argv[1], 'r')
elif len(sys.argv) <2:
    infile = raw_input("Name of file: ")
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
data = open(infile, 'r')
for x in data:
    x = float(x) # HERE IS THE PROBLEM
    numLst.append(x.strip('\n').split(" "))
 
a = array(numLst)
print a
 

Input file looks like this:

1.0 2.0
2.0 3.5
2.5 5.0
3.5 6.5

which are data points, x and y respectively.

Thanks!

TMS

Not sure about your error message but if sys.argv == 2 then wouldn't you be opening the file twice? Probably nothing to do with the problem but just thought I'd mention it.

Also, you didn't read the file. Try:

data = infile.read()
 

I think that might work.

[TMS]

ok, I see now. But I'm not sure how to use that with my for loop. Or do I use it in the for loop? Shouldn't I convert the file data to float before I append it to my list?

[TMS]

Not sure about your error message but if sys.argv == 2 then wouldn't you be opening the file twice?

Also, you didn't read the file. Try:

data = infile.read()
 

I think that might work.

No.
2 = the second argument in the file. Note the 'len' before sys.argv

It reads the file fine. That isn't the problem. I'm trying to convert the data from strings to floats.

[ilikepython]

No.
2 = the second argument in the file. Note the 'len' before sys.argv

It reads the file fine. That isn't the problem. I'm trying to convert the data from strings to floats.

Oh sorry, but for the file opening I meant

 
if len(sys.argv) == 2:
    infile = open(sys.argv[1], 'r')
elif len(sys.argv) <2:
    infile = raw_input("Name of file: ")
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
 data = open(infile, 'r')
 

You have two open calls. I'm sorry if I'm being stupid but it doesn't really matter so forget about it.

[TMS]

you aren't being stupid.
My assignment requires that only one file is provided on the command line.

Do you have a better way of doing it?

In tests what I've done works fine. If I provide more than one file I get an error, which is what I want.

Perhaps I'm missing something here. The first arg is 0, second is 1, etc. If the first arg is the program name, the second is the file name I want opened. So, the length of the args should equal 2 or there are two many args, right?

[bvdet]

ok, I see now. But I'm not sure how to use that with my for loop. Or do I use it in the for loop? Shouldn't I convert the file data to float before I append it to my list?

>>> numList = []
>>> fList = open('data_file').readlines()
>>> for x in fList:
...     numList.append([float(i) for i in x.strip().split()])
...     
>>> numList
[[1.0, 2.0], [2.0, 3.5], [2.5, 5.0], [3.5, 6.5]]
>>> 

If you create a file object, don't forget to close it. Done like above, no file object is created.

[ilikepython]

you aren't being stupid.
My assignment requires that only one file is provided on the command line.

Do you have a better way of doing it?

In tests what I've done works fine. If I provide more than one file I get an error, which is what I want.

Perhaps I'm missing something here. The first arg is 0, second is 1, etc. If the first arg is the program name, the second is the file name I want opened. So, the length of the args should equal 2 or there are two many args, right?

Ok, sys.argv[1] is the file name. Right? If that's the case then I believe you should get an error message. I tried it:

>>> infile = open("C:\\testfile.txt", "r")
>>> data = open(infile, "r")
 
Traceback (most recent call last):
  File "<pyshell#5>", line 1, in <module>
    data = open(infile, "r")
TypeError: coercing to Unicode: need string or buffer, file found
>>> 
 

Did it work for you?
Please correct me if I am wrong?

[ghostdog74]

something you might want to take a look

....
if len(sys.argv) == 2:
    data = open(sys.argv[1], 'r')
elif len(sys.argv) <2:
    infile = raw_input("Name of file: ")
        data = open(infile, 'r')
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
for x in data:
....
 

sys.argv[0] is the program name, argv[1] is the first argument and so on.

[bvdet]

ok, I see now. But I'm not sure how to use that with my for loop. Or do I use it in the for loop? Shouldn't I convert the file data to float before I append it to my list?

Here's a one-liner:

>>> [[float(x),float(y)] for x,y in [j.split() for j in open('data.txt').read().split('\n')]]
[[1.0, 2.0], [2.0, 3.5], [2.5, 5.0], [3.5, 6.5]]
>>>

[TMS]

YEAH!!! thank you. Now I can get down to the complexities of linear regression, which is more busy work than anything. Thank you!

This is what it looks like right now, there are no error messages:

#! /usr/bin/env python
import sys
import os
from numpy import * 
numLst = [] #an empty list
if len(sys.argv) == 2:
    infile = open(sys.argv[1], 'r')
    for x in infile:
        numLst.append([float(i) for i in x.strip().split()]) 
elif len(sys.argv) <2:
    infilename = raw_input("Name of file: ")
    data = open(infilename, 'r')
    for x in data:
        numLst.append([float(i) for i in x.strip().split()])
elif len(sys.argv) > 2:
    raise SyntaxError, "Too many arguments"
a = array(numLst)
print a
b = a[0:, 1] #this is just me manipulating the list a bit
print b
c = a[0:,1]
print
bA = b[:1]
g = bA
print numLst
 
 

>>> numList = []
>>> fList = open('data_file').readlines()
>>> for x in fList:
...     numList.append([float(i) for i in x.strip().split()])
...     
>>> numList
[[1.0, 2.0], [2.0, 3.5], [2.5, 5.0], [3.5, 6.5]]
>>> 

If you create a file object, don't forget to close it. Done like above, no file object is created.