用二进制值替换CSV文件中的分类值

Question

I have a clinical data set and I have to replace

• the 1st column values 'DECEASED' with 1, if the value 'Date' > 365 else replace with 0 (zero),
• the value 'LIVING' with one if 'Day_to_follow_up' > 365

In addition I need to assign the ages

• 0-25 to bin 0,
• 25-50 to bin 1,
• 50-75 to bin 2
• above 75 to bin 4.

Here is my code.

import csv
import pandas as pd
with open('combined_file', 'rb') as f,open('newFile', 'wb') as out:
    reader = csv.reader(f)


    writer = csv.writer(out)
    for row in reader:
        #print "AABB"
        if 'DECEASED' in row[1]:
            if row[10]>365:
                row[1]=1
                writer.writerow(row)
            elif row[10]<365:
                row[1]=0
                writer.writerow(row)
        if 'LIVING' in row[1]:
            if row[11]>365:
                row[1]=1
                writer.writerow(row)

sample input

sample id , status , age ,gender ,date ,days_to_last_followup
0     ,    Deceased , 42 , M  ,   326 ,    149
1     ,    Deceased , 56 , F  ,   500 ,    30
2     ,    living   , 43 ,M   ,   25  ,    150

sample output

sample id , status , age ,gender,date ,days_to_last_followup
0     ,       0    , 1 ,  M    ,326 ,    149
1     ,       1    , 2 , F     ,500 ,    30
2     ,       0    , 1 ,M   ,   25  ,    150

Answer 1

I'm not sure what your question is, based off this post. Either way, the logical structure would have an issue if both 'Deceased' and 'Living' were in row[1]. I'd suggest you create some test cases to look for bad data, since ETL processes routinely have to deal with unexpected data formats/fields.

I'm also not sure why you are importing the pandas library. You don't seem to be calling it anywhere in the code you posted.

Answer 2

Your code is a good starting point - a few things that the code does not cover:

• What happens when 'DECEASED' and 'LIVING' are both in row[1] ? Your code will write two rows. To fix this, set the if 'LIVING' to elif 'LIVING' .
• You need an else case to catch what happens when neither DECEASED or LIVING is in row[1] .