How to extract column and row in csv using python

import csv

def main():
    with open('file.csv', 'rb') as inf:
        data = [(int(row['rainfall']), row['']) for row in csv.DictReader(inf)]

    data.sort()
    print data[0]

if __name__=="__main__":
    main()

returns

(300, 'Missouri')

One way to do this would be to use the csv module's DictReader class to write a function to extract the column of data. DictReader will take care of handling the first row of field names automatically. The built-in min() function can then be used to determine the item with the smallest value in the column.

import csv

def csv_extract_col(csvinput, colname, key):
    """ extract a named column from a csv stream into a dictionary
          colname:  name of columm to extract
          key:  name of another columm to use as keys in returned dict
    """
    col = {}
    for row in csv.DictReader(csvinput):
        col[row[key]] = row[colname]
    return col

if __name__=='__main__':
    import StringIO

    csvdata = """\
"","min","max","rainfall","days_clear"  # field name row
"Missouri",-2,10,300,23
"Amsterdam",-3,5,1212,34
"LA",10,20,1000,54
"""
    csvfile = StringIO.StringIO(csvdata)

    rainfall = csv_extract_col(csvfile, 'rainfall', '')
    print rainfall
    # {'Amsterdam': '1212', 'LA': '1000', 'Missouri': '300'}

    print min(rainfall.iteritems(), key=lambda r: float(r[1]))
    # ('Missouri', '300')

import StringIO
import csv

example = """"","min","max","rainfall","days_clear"
"Missouri",-2,10,300,23
"Amsterdam",-3,5,1212,34
"LA",10,20,1000,54
"""

data_in = StringIO.StringIO(example)
#data_in = open('mycsvdata.csv')

def read_data(data_in):
  reader = csv.reader(data_in)
  cols = []
  results = {}
  for row in reader:
    if not cols:
      cols = row
      continue
    row = [ int(x) if x.lstrip('-').isdigit() else x for x in row ]
    results[row[0]] = dict(zip(cols[1:],row[1:]))
  return results

data = read_data(data_in)

min(data.items(),key=lambda x: x[1].get('rainfall'))

Returns

('Missouri', {'max': 10, 'days_clear': 23, 'rainfall': 300, 'min': -2})

To read from a file, you need to remove all code that deals with a string:

   reader = csv.reader(open('file.csv', 'rb'))
   rainfall = csv_extract_col(reader, 'rainfall', '')

Update: Sorry, it neads a bit more work than that. The first arg of csv_extract_col will be used as the first arg of csv.DictReader so (in this case) it should be an open file object, and should never be a csv.reader instance. See below:

import csv

### def csv_extract_col(csvinput, colname, key):
### exactly as provided by @martineau

if __name__ == '__main__':
    import sys
    filename, data_col_name, key_col_name = sys.argv[1:4]
    input_file_object = open(filename, 'rb')
    result_dict = csv_extract_col(input_file_object, data_col_name, key_col_name)
    print result_dict
    print min(result_dict.iteritems(), key=lambda r: float(r[1]))

Results:

command-prompt>\python27\python joj_csv.py joj.csv rainfall ""
{'Amsterdam': '1212', 'LA': '1000', 'Missouri': '300'}
('Missouri', '300')

command-prompt>\python27\python joj_csv.py joj.csv days_clear ""
{'Amsterdam': '34', 'LA': '54', 'Missouri': '23'}
('Missouri', '23')

Update 2 in response to comment """there must be something i missed out.. i tried.. [what looks like @martineau's function] with the above main function you define. Then in my shell, i define python rainfall "". But it gives me KeyError: 'rainfall'"""

Two possibilities:

(1) You made a mistake patching the pieces of source code together. Check your work.

(2) Your file doesn't have the expected heading row contents. Try some debugging e.g. change @martineau's code so that you can insert a print statement etc. to show what the csv.DictReader thinks about your heading row:

reader = csv.DictReader(csvinput)
print "fieldnames", reader.fieldnames
assert colname in reader.fieldnames
assert key in reader.fieldnames
for row in reader:

If you are still stuck, show us ALL of your code plus the full traceback and error message -- either edit your question or put it up on pastbin or dropbox; DON'T put it into a comment!!

My code for cases in which there are several cities having the same minimum or several cities having the same maximum:

import csv

def minmax_col(filename,key,colname):
    with open(filename,'rb') as csvfile:
        rid = csv.DictReader(csvfile,
                             fieldnames=None,
                             quoting=csv.QUOTE_NONNUMERIC)

        mini = float('inf')
        maxi = float('-inf')
        limin = limax =[]

        for row in rid:
            if row[colname] == maxi:
                limax.append(row[key])
            elif row[colname] > maxi:
                maxi = row[colname]
                limax = [row[key]]

            if row[colname] == mini:
                limin.append(row[key])
            elif row[colname] < mini:
                mini = row[colname]
                limin = [row[key]]

    return (key,(maxi,limax),(mini,limin))



key = 'rainfall'
city,(Ma,liMa),(mi,limi) = minmax_col('filename.csv','',key)
print 'Cities analysed on ' + repr(key) + ' parameter :'
print 'maximum==',Ma,'  cities :',', '.join(liMa)
print 'minimum==',mi,'  cities :',', '.join(limi)

print 

key = 'min'
city,(Ma,liMa),(mi,limi) = minmax_col('filename.csv','',key)
print 'Cities analysed on ' + repr(key) + ' parameter :'
print 'maximum==',Ma,'  cities :',', '.join(liMa)
print 'minimum==',mi,'  cities :',', '.join(limi)

On a file like that:

"","min","max","rainfall","days_clear"
"Missouri",-2,10,300,23
"Amsterdam",-3,5,1212,34
"Oslo",-2,8,800,12
"LA",10,20,1000,54
"Kologoro",28,45,1212,1

the result is

Cities analysed according the 'rainfall' parameter :
maximum== 1212.0   cities : Amsterdam, Kologoro
minimum== 300.0   cities : Missouri

Cities analysed according the 'min' parameter :
maximum== 28.0   cities : Kologoro
minimum== -3.0   cities : Amsterdam