There's a nice NumPy-based Python package called Tabular (http://www.parsemydata.com/tabular/index.html http://www.parsemydata.com/tabular/index.html) that make

There's a nice NumPy-based Python package called Tabular (http://www.parsemydata.com/tabular/index.html http://www.parsemydata.com/tabular/index.html) that makes this super easy: import numpy as np import tabular # CSV with Region, City and Sales columns data = tabular.tabarray(SVfile = 'data.csv') # Calculate the total sales within each region summary = data.aggregate(On = ['Region'], AggFuncDict = {'Sales':np.sum}, AggFunc = len) summary.saveSV('summary.csv')