For this assignment we are required to demonstrate the use of Python, databases and Git for version control.

The request is that we are to:

Suppose you have been hired by a renewable fuels company, who want to locate a processing plant somewhere in Ireland. They already have several raw materials locations in Ireland, where they cut down trees. They have decided that the new plant will be at one of those locations. The idea is that wood from all locations will be transported by road to the processing plant, and then the product will be transported by road to some port. You have been asked to choose the location for the processing plant and the choice of port, so as to minimize the total road transportation costs.

For this we have been provided a database with two tables 'ports' and 'location' . We have to read the data out of them, store them in appropriate Python data structures. Then use a brute force (try all combinations) algorithm, which will perform the necessary calculations and provide the output for our business stakeholders.

My approach is well commented throughut the respective scripts themselves. But effectively location_planning.py is a script which reads the databases into 'Location' objects with sensibly named parameters. These Location objects are of two types 'Port' and 'Plant' which are then stored in their own respective lists .

These lists are then compared by the rank_pairs function which does the brute force comparison of every 'Port' and every 'Plant'. It stores its output as a list of named tuples. These named tuples have a name field I constructed for the plant and port from their type and coordinates as they lacked a true unique identifier or key column. The value it stores as the distance is the euclidean distance in the same units as the coordinates themselves. The named tuple also stores the potential production volume possible from that Plant This list is then sorted by ascending distance and converted to a pandas dataframe object.

This dataframe object is then printed to the terminal. However as I know most business stakeholders prefer using csv format. So I also write out to a csv 'output.csv' for ease of communication. This script also has associated unit tests in 'location_planning_test.py' These tests allow us to make sure all the code is working as intended. This will also make it easier to extend or further develop this code without introducing new bugs.

All the code is written mostly in line with the Google Python style guide. So I have used 2 space indentations throughout.

Launching location_planning.py is done as follows and should return something similar to this:

python location_planning.py
                       Plant                     Port  Distance  Production
0   Plant Lon:52.34 Lat:6.48  Port Lon:52.27 Lat:6.39  0.114018      298978
1   Plant Lon:52.86 Lat:8.99   Port Lon:52.7 Lat:8.63  0.393954       89499
2   Plant Lon:53.22 Lat:6.68  Port Lon:53.33 Lat:6.25  0.443847      200068
3   Plant Lon:53.66 Lat:6.69  Port Lon:53.33 Lat:6.25  0.550000      210817
4    Plant Lon:53.18 Lat:6.8  Port Lon:53.33 Lat:6.25  0.570088      163358
5   Plant Lon:52.84 Lat:6.92  Port Lon:52.27 Lat:6.39  0.778332      159794
6   Plant Lon:52.84 Lat:6.92  Port Lon:53.33 Lat:6.25  0.830060      159794
7   Plant Lon:52.66 Lat:7.26  Port Lon:52.27 Lat:6.39  0.953415      259696
8   Plant Lon:52.36 Lat:7.71   Port Lon:52.7 Lat:8.63  0.980816       60573
...

The unit tests location_planning_test.py can be run as follows and should return similar to this:

python location_planning_test.py
........
----------------------------------------------------------------------
Ran 8 tests in 0.001s

In this problem we have limited information, so the business action is not certain. Depending on other factors like the cost of each location and the production volumes required we may need one or more sites to cover our required volumes. But deciding on the information we have alone here I would recommend investing in the Plant at 'Lon:52.34 Lat:6.48' and the Port at 'Lon:52.27 Lat:6.39'. This minimizes the euclidean distance to '0.114018' as mentioned above. While the potential production volume is the best of any Plant at '298978' tons per year.

sqlite3

pandas

enum 'https://pypi.python.org/pypi/enum34/'

unittest

mock