What's the easiest way of getting this data into a Pandas Dataframe?

Question

I came across this dataset:

http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/auto-mpg.data

and I couldn't find a simple way of getting this into a Pandas Dataframe. I manually parsed this into a list of lists and then called the Dataframe constructor, but is there an easier way of doing this. Thanks!

Answer 1

Try using pandas.read_fwf and specify a list of column widths (including whitespace):

In [35]: url = 'http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/auto-mpg.data'

In [36]: widths = [7, 4, 10, 10, 11, 7, 4, 4, 30]

In [37]: df = pd.read_fwf(url, widths=widths, header=None, na_values=['?'])

In [38]: df.irow(0)
Out[38]: 
X0                              18
X1                               8
X2                             307
X3                             130
X4                            3504
X5                              12
X6                              70
X7                               1
X8    "chevrolet chevelle malibu"

Name: 0