将不同字典中相同键的多个值相加 Python

Question

I have a filepath full of CSV files. I am using Python glob to open them and csv.DictReader() to read through them and parse the data into dictionaries with the headers as keys.

The data in the CSV files looks like this:

CSVfile1:

Name,A,B,C,D,Date
John,-1,2,4.0,-5.1,3/23/2016
Jacob,0,3,2.0,-2.3,3/23/2016
Jinglehimmer,1,100,5.0,-.1,3/23/2016

CSVfile2:

Name,A,B,C,D,Date
John,5,4,1.0,-1,3/24/2016
Jacob,0,1,7.0,-.1,3/24/2016
Schmidt,10,9,8,7,3/24/2016

I am trying to SUM the data in the A , B , C and D columns for each name over a set date period (say the past 2 days). For example, I am trying to get a new list of dictionaries that looks like this:

{Name: John, A: 4, B: 6, C: 5.0, D: -6.1, Date: 2}
{Name: Jacob, A: 0, B: 4, C: 9.0, D: -2.4, Date: 2}
{Name: Jinglehimmer, etc.}
{Name: Schmidt, etc.}

Here is the code I have so far that I know works. This opens each CSV and creates a dictionary for each row and allows me to iterate through the dictionaries:

import csv
import glob

path = "."

newdict = {}

for filename in glob.glob(path):
    with open(filename) as csv_file:
        for row in csv.DictReader(csv_file):

Edit: I tried simply summing all the key values into a new dictionary, but I run into an int+str error.

for k in row.keys():
    newdict[k] = newdict.get(k,0) + row[k]

I am also not sure how to filter by the Date: key to only get x days of data.

Any help or points in the right direction are much appreciated.

Answer 1

The following approach should work:

import csv
import glob
from datetime import datetime, timedelta, date


days = 2
since = datetime.combine(date.today(), datetime.min.time()) - timedelta(days = days)
required_fields = ['A', 'B', 'C', 'D']

path = "."
newdict = {}

output = {}

for filename in glob.glob(path):
    with open(filename) as csv_file:
        for row in csv.DictReader(csv_file):
            if datetime.strptime(row['Date'], '%m/%d/%Y') >= since:
                name = row['Name']

                try:
                    cur_entry = output[name]
                    entry = {field : cur_entry[field] + float(row[field]) for field in required_fields}
                except KeyError as e:
                    entry = {field : float(row[field]) for field in required_fields}
                    entry['Date'] = days

                output[name] = entry

for name, entry in output.items():                
    print name, entry

Which for the data you have given will display:

Jacob {'A': 0.0, 'C': 9.0, 'B': 4.0, 'D': -2.4}
Jinglehimmer {'A': 1.0, 'Date': 2, 'C': 5.0, 'B': 100.0, 'D': -0.1}
John {'A': 4.0, 'C': 5.0, 'B': 6.0, 'D': -6.1}
Schmidt {'A': 10.0, 'Date': 2, 'C': 8.0, 'B': 9.0, 'D': 7.0}

A datetime object can be used to help with measuring the time interval.