How to run gradient_descent.py?

[fujishoichi]

Hi,

I tried running gradient_descent.py.
But it failed.

$ python gradient_descent.py 
Traceback (most recent call last):
  File "gradient_descent.py", line 73, in <module>
    x, y = np.loadtxt("data.txt", delimiter= "\t", unpack = True)
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1146, in loadtxt
    for x in read_data(_loadtxt_chunksize):
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1074, in read_data
    items = [conv(val) for (conv, val) in zip(converters, vals)]
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1074, in <listcomp>
    items = [conv(val) for (conv, val) in zip(converters, vals)]
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 781, in floatconv
    return float(x)
ValueError: could not convert string to float: 'The dataset for gradient descent example can be downloaded from: http://themlbook.com/wiki/doku.php?id=gradient_descent'

I downloaded the CSV pointed in data.txt and loaded it instead of it, but it didn't work as well:

$ python gradient_descent.py
Traceback (most recent call last):
  File "gradient_descent.py", line 73, in <module>
    x, y = np.loadtxt("Advertising.csv", delimiter= "\t", unpack = True)
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1146, in loadtxt
    for x in read_data(_loadtxt_chunksize):
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1074, in read_data
    items = [conv(val) for (conv, val) in zip(converters, vals)]
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 1074, in <listcomp>
    items = [conv(val) for (conv, val) in zip(converters, vals)]
  File "/Users/fujisho/.pyenv/versions/3.7.4/lib/python3.7/site-packages/numpy/lib/npyio.py", line 781, in floatconv
    return float(x)
ValueError: could not convert string to float: ',TV,radio,newspaper,sales'

It would be helpful if you can provide any tips to solve the problem.

[Qu3tzal]

Looks like the gradient_descent.py script expects a tabular CSV.
You have to remove the header (the first line of Advertising.csv), remove the ids (the first column of Advertising.csv), and then replace the comas by tabs (real tabs, not 2/4/8 spaces).

[dluvishis]

1. save advertising.csv to the same folder with the original file
2. create a new folder "Illustrations" within the same folder
3. replace lines 12 and 72 with:
   x, y = np.loadtxt("advertising.csv", delimiter= ",", skiprows=1, usecols=(2,4), unpack = True)
   to ignore the header, use "," instead of tabular and use specific columns instead of all 5

It will work now.