Pandas csv-import: Keep leading zeros in a column

问题:

I am importing study data into a Pandas data frame using read_csv.

My subject codes are 6 numbers coding, among others, the day of birth. For some of my subjects this results in a code with a leading zero (e.g. "010816").

When I import into Pandas, the leading zero is stripped of and the column is formatted as int64.

Is there a way to import this column unchanged maybe as a string?

I tried using a custom converter for the column, but it does not work - it seems as if the custom conversion takes place before Pandas converts to int.

回答1:

As indicated in this question/answer by Lev Landau, there could be a simple solution to use converters option for a certain column in read_csv function.

converters={'column_name': lambda x: str(x)}

You can refer to more options of read_csv funtion in pandas.io.parsers.read_csv documentation.

Lets say I have csv file projects.csv like below:

project_name,project_id
Some Project,000245
Another Project,000478

As for example below code is triming leading zeros:

import csv
from pandas import read_csv

dataframe = read_csv('projects.csv')
print dataframe

Result:

me@ubuntu:~$ python test_dataframe.py 
      project_name  project_id
0     Some Project         245
1  Another Project         478
me@ubuntu:~$

Solution code example:

import csv
from pandas import read_csv

dataframe = read_csv('projects.csv', converters={'project_id': lambda x: str(x)})
print dataframe

Required result:

me@ubuntu:~$ python test_dataframe.py 
      project_name project_id
0     Some Project     000245
1  Another Project     000478
me@ubuntu:~$

回答2:

here is a shorter, robust and fully working solution:

simply define a mapping (dictionary) between variable names and desired data type:

dtype_dic= {'subject_id': str, 
            'subject_number' : 'float'}

use that mapping with pd.read_csv():

df = pd.read_csv(yourdata, dtype = dtype_dic)

et voila!

回答3:

I don't think you can specify a column type the way you want (if there haven't been changes reciently and if the 6 digit number is not a date that you can convert to datetime). You could try using np.genfromtxt() and create the DataFrame from there.

EDIT: Take a look at Wes Mckinney's blog, there might be something for you. It seems to be that there is a new parser from pandas 0.10 coming in November.

回答4:

If you have a lot of columns and you don't know which ones contain leading zeros that might be missed, or you might just need to automate your code. You can do the following:

df = pd.read_csv("your_file.csv", nrows=1) # Just take the first row to extract the columns' names
col_str_dic = {column:str for column in list(df)}
df = pd.read_csv("your_file.csv", dtype=col_str_dic) # Now you can read the compete file

You could also do:

df = pd.read_csv("your_file.csv", dtype=str)

By doing this you will have all your columns as strings and you won't lose any leading zeros.