CODEWITHSUNDEEP
pythonarraysnumpypandasscikit-learn
renno
renno

Reputation: 2827

Creating NumPy array with Pandas

I'm trying to use scikit with some data that are in one spreadsheet (.xlsx). To achieve this I'm using Pandas to read the spreadsheet and then I was going to use numpy to use scikit.

The problem here is that when I covert my DF structure to numpy, I lose almost all the data! I think it is because it doesn't have columns names. There are just raw data. EX:

28.7967 16.0021 2.6449 0.3918 0.1982

31.6036 11.7235 2.5185 0.5303 0.3773

162.052 136.031 4.0612 0.0374 0.0187

My code so far:

def split_data():
    test_data = pd.read_excel('magic04.xlsx', sheetname=0, skip_footer=16020)
    #code below prints correctly the data
    print test_data.iloc[:, 0:10] 

    #none of the code below work as expected 
    test1 = np.array(test_data.iloc[:, 0:10])
    test2 = test_data.as_matrix()

I'm really lost here. Any help would be very welcome...

Upvotes: 2

Views: 86

Answers (1)

Ami Tavory
Ami Tavory

Reputation: 76297

I'd suggest that you use header=None in read_excel. See the following:

df = pd.read_excel('stuff.xlsx')
>> df
    28.7967 16.0021 2.6449  0.3918  0.1982
0   31.6036 11.7235 2.5185  0.5303  0.3773
1   162.0520    136.0310    4.0612  0.0374  0.0187

>> df.ix[:, 1: 2]

0
1

Versus:

df = pd.read_excel('stuff.xlsx', header=None)
>> df

0   1   2   3   4
0   28.7967 16.0021 2.6449  0.3918  0.1982
1   31.6036 11.7235 2.5185  0.5303  0.3773
2   162.0520    136.0310    4.0612  0.0374  0.0187

>> df.ix[:, 1: 2]
    1   2
0   16.0021 2.6449
1   11.7235 2.5185
2   136.0310    4.0612

Upvotes: 2

Related Questions