Carpool ***

Some friends have organized carpool them-self.
At the beginning of each accounting period, they write down the numbers of the car's odometer.
Every time somebody drives the car, she/he writes down the numbers when the car is returned and hers/his name.

carpool.dat

17000  start
17100  Daniel
17220  Sara
17310  David
17410  Daniel
17550  Sara
17800  David

Try to write an awk script that calculates what distance has everyone traveled with the car. Hints: Differences between numbers in two consecutive lines might be a good start. Associative arrays might come handy.

Answer for the data above:

           Sara: 260
         Daniel: 200
          David: 340

The problem is rather easy to solve with awk (it is my humble opinion). Can it be solved easier with some of your favorite tools? (I am interested of alternative approaches)

Possible solution:

awk '{dist[$2]=dist[$2]+($1-prev); prev=$1} END{ for (i in dist) if (i !="start") printf("%15s: %g\n",i,dist[i]) }' carpool.dat

solution in Python (credits to Mihai Croicu)

import pandas as pd
frame = pd.read_csv('odometer',sep='  ',header=None)                                                                                         
frame['driven']=frame[0]-frame[0].shift(1)                                                                                                                          
frame.groupby([1])['driven'].sum()