Print all rows related to minimum values from another column based on distinct values of a specific column from csv file using python csv

Question

I have a CSV file with following structure:

Id,User,P_Name,P_Code,Rate

1,U1,P1,1234,21.5

2,U1,P2,7483,20

3,U1,P3,8945,29.5

4,U2,P1,1234,80

5,U2,P2,7483,23.5

6,U2,P3,8945,30

7,U3,P1,1234,15

8,U3,P2,7483,27.3

9,U3,P3,8945,,29.7

I want to print complete rows for minimum value of each product. For instance, here it would be:

7,U3,P1,1234,15

2,U1,P2,7483,20

3,U1,P3,8945,29.5

I am new to python and unable to proceed after this:

import csv
with open('sample.csv', 'rb') as csvfile:
        filereader = csv.reader(csvfile, delimiter=',', quotechar='|')
        headers=next(filereader)
        data = []
        for row in filereader:
                data.append(row[2])
        print (data)

Here, I am getting a list of P_Name values and unable to figure out how to get minimum value of each distinct product.

Answer 1

First append the whole CVS row, not just the third items of the row (like row[2] )

import csv
with open('sample.csv', 'rb') as csvfile:
    filereader = csv.reader(csvfile, delimiter=',', quotechar='|')
    headers=next(filereader)
    data = []
    for row in filereader:
            data.append(row)
    print (data)

Then build a dict that use P_name as a key, with the whole row as value. The dicts thus stores the whole row, with row[2] as key. Then iter over each row, replace the current dict value with a new, if a lower price is found.

filter = {}
for item in data:
   if item[2] not in filter.keys():     #First if dict already has an entry in dict
           filter[item[2]] = item       #if no entry ad entry
   elif item[4] < filter[item[2]][4]:   #if entry compare between entry in dicts and cvs line. 
                                        #Both refer to [4] so booth compare the rate of the CVS column
           filter[item[2]] = item

And to print your values.

 for item in filter.keys():
      print item,' : ',filter[item]

Per your second remarks, then is it better to add extra info to the values. You could opt for a list that holds data for price in index 0 price[0] and users in index 1, price[1]

filter = {}
for item in data:
   if item[2] not in filter.keys():     #First if dict already has an entry in dict
           filter[item[2]] = [item[4], [item[1]]       #if no entry ad entry, the dict value is a list.
 #Filter Dict Value explained ..
 #Index 0 stores the the price of the product
 #Index 1 stores a list of users that have the product at this value. 

   elif   item[4] == filer[item[2]][0]:                #price is identical add another user to the dict[product][second slot of list]
              filter[item[2]][1].append(item[1])       #filter[productCode][second index] APPEND [New user ]



   elif item[4] < filter[item[2]][0]:   

#If a lower product rate has been found, then reset the value of the dict. 
#And store new lower price, with it's corresponding user.                                       
           filter[item[2]] = [item[4], [item[1]]

Answer 2

Thanks for the response. I had modified your code a bit to make it a bit simpler.

filter = {} for item in data: if item[2] not in filter.keys():
filter[item[2]] = item
elif item[4] == filter[item[2]][4]:
filter[item[2]].append(item) elif item[4] < filter[item[2]][4]:
filter[item[2]] = item

Although, it works fine. However, I am facing some issues with formatting of the result after updating row 5 (post header) in the csv file from

5,U2,P2,7483,23.5 

to

5,U2,P2,7483,20

And then printed results using following code:

 for item in filter.keys():
                print filter[item]

Results are as follows:

['2', 'U1', 'P2', '7483', '20', ['5', 'U2', 'P2', '7483', '20']]
['3', 'U1', 'P3', '8945', '29.5']
['7', 'U3', 'P1', '1234', '15']

whereas if in case there are two users paying identical price for a particular product then instead of appending those details along with previous user, I want to display it as a separate entry and in similar format as csv file (without brackets and quotes), like:

2,U1,P2,7483,20 
5,U2,P2,7483,20
3,U1,P3,8945,29.5
7,U3,P1,1234,15