Use pd.read_csv(), specified data as np.float32, but got string

Question

When I tried to read the data from a csv file. I want to set the 'Data' column's datatype as float, so my code is like:

data = pd.read_csv('data.csv', index_col=0, 
                   parse_dates=[0], 
                   skiprows=[0], names=['Data'],
                   dtype={'Data':np.float32})

But the result I got is still list of strings. Is there any reason and any solution for it?

[5]: data.values

[5]: array([['1188.0'],
            ['1377.0'],
            ['1279.0'],
            ['1461.0'],
            ['1146.0'],
            ['1287.0'],
            ['1259.0']], dtype=object)

jezrael · Accepted Answer · 2019-11-14 07:17:03Z

1

Add value to names parameter:

import pandas as pd
import numpy as np
from io import StringIO

temp="""2019-01-02,1188.0
2019-01-03,1377.0
2019-01-04,1279.0
2019-01-05,1461.0
2019-01-06,1146.0
2019-01-07,1287.0
2019-01-08,1259.0"""
#after testing replace 'pd.compat.StringIO(temp)' to 'data.csv'
df = pd.read_csv(StringIO(temp),
                 index_col=0,
                 parse_dates=[0],
                 names=['dates','Data'],
                 dtype={'Data':np.float32})

print (df)
              Data
dates             
2019-01-02  1188.0
2019-01-03  1377.0
2019-01-04  1279.0
2019-01-05  1461.0
2019-01-06  1146.0
2019-01-07  1287.0
2019-01-08  1259.0

print (df.dtypes)
Data    float32
dtype: object

print (df.index)
DatetimeIndex(['2019-01-02', '2019-01-03', '2019-01-04', '2019-01-05',
               '2019-01-06', '2019-01-07', '2019-01-08'],
              dtype='datetime64[ns]', name='dates', freq=None)

edited Nov 14, 2019 at 7:17

answered Nov 14, 2019 at 5:54

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Kid Over a year ago

Thanks for your answer, but actually the csv file includes 2 columns: date and data. If I remove the parse_date, I can have the correct datatype. So does it means that I couldn't achieve both in one line? Does the parse_date function conflicts with the datatype specification? Thanks!

jezrael Over a year ago

@Kid - So I think you need df = pd.read_csv('data.csv', index_col=0, names=['Data'], dtype={'Data':np.float32})

jezrael Over a year ago

@Kid - Solution was changed completely.

Collectives™ on Stack Overflow

Use pd.read_csv(), specified data as np.float32, but got string

1 Answer 1

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related