Skip to content


Folders and files

Last commit message
Last commit date

Latest commit


Repository files navigation


Welcome to EnergyDataSimulationChallenge!

This project allows applicants to demonstrate their ability to analyze and develop software that makes use of big energy production data. Please complete one (or more) of the several challenges we have prepared. Pull-requests are also welcome.



  1. Fork this repository
  2. Create a new branch (please name the branch challengeX/YOURNAME (eg. challenge1/shirakia))
  3. Create your development directory inside the analysis/ or webapp/ directory
  4. Design, write and commit your program to the above branch
  5. Push the branch
  6. Create a Pull Request


  • Avoid working on any branch except your own branch
  • Avoid committing files other than those in your own directory


  • For Full-Time Employment
    • A chance at a final interview with the CTO and engineers.
  • For Paid Internship
    • A paid internship offer (3 month internship programme with competitive salary)
    • Includes FREE accommodation (private room) and own desk at coworking space in London if you wish to work with us here.
    • Optionally, you may work remotely from a location of your choice.
    • After successful 3 month programme, you may be offered an extension, or another position (permanent, longer contract, additional intern, etc..)

Challenge 1 - Energy Production Data Simulation

We have prepared energy production data for 500 houses. For each house, there is monthly data from July, 2011 to June, 2013. The data contains temperature and daylight data.

Please make a model for predicting EnergyProduction using data from July 2011 to May 2013. On that basis, predict EnergyProduction on June 2013 for each house, and calculate MAPE (Mean Absolute Percentage Error). You can use any algorithm, including multiple-variables regression, polynomial regression, Neural network, SVM, etc...

We will check the following:

  • accuracy of prediction (MAPE)
  • algorithm choice
  • parameter tuning
  • programming skill

Please make sure to consider all of the above criteria (not just accuracy of prediction) when completing each challenge.


The input dataset file inside the data/ directory has the following format:

$ head data/dataset_500.csv | column -s, -t
ID  Label  House  Year  Month  Temperature  Daylight  EnergyProduction
0   0      1      2011  7      26.2         178.9     740
1   1      1      2011  8      25.8         169.7     731
2   2      1      2011  9      22.8         170.2     694
3   3      1      2011  10     16.4         169.1     688
4   4      1      2011  11     11.4         169.1     650
5   5      1      2011  12     4.2          199.5     763
6   6      1      2012  1      1.8          203.1     765
7   7      1      2012  2      2.8          178.2     706
8   8      1      2012  3      6.7          172.7     788

The first line of the file gives the format name. The rest of the file describes EnergyProduction data for 500 houses. Each dataset consists of 24 lines showing monthly temperature and daylight EnergyProduction data.

training_dataset_500.csv and test_dataset_500.csv are subsets of dataset_500.csv. test_dataset_500.csv includes only June 2013 data of each house (the rest can be found in training_dataset_500.csv).

You can use any of the given data you like; but please do not forget that you can use only data from July 2011 to May 2013 for training.


Output is predicted_energy_production.csv, mape.txt and other files. Please place these files in challenge1/analysis/YOURNAME/.

  1. predicted_energy_production.csv Must include House column and EnergyProduction column for each line. Any csv file that we can find which columns means House and EnergyProduction is also acceptable.
  2. mape.txt Need to include just MAPE value. Minimize it.
  3. another files Should include files you use, edit or write -- like R source code, batch Python file, excel file, etc.. These files will help us understand your thought process. You are not required to commit any files that contain sensitive information.

Challenge 2 - Visualization of Energy Consumptions

The following task is intended to give us an idea of your data visualisation skills. Please use the tools and programming language with which you are most familiar.


  1. Download the data-set total-watt.csv
  2. The data-set consists of two columns: a time stamp and the energy consumption
  3. visualise the data-set
  4. visualise the data-set as values per day
  5. cluster the values per day into 3 groups: low, medium, and high energy consumption
  6. visualise the clusters (How you visualize the data is up to you. Please show us your imagination and creativity!)


dataset file is in data/ directory as follows.

$ head data/total_watt.csv| column -s, -t
2011-04-18 13:22:00  925.840613752523
2011-04-18 13:52:00  483.295891812865
2011-04-18 14:22:00  915.761633660131
2011-04-18 14:52:00  609.043490935672
2011-04-18 15:22:00  745.155434458509
2011-04-18 15:52:00  409.855947368421
2011-04-18 16:22:00  434.084038321073
2011-04-18 16:52:00  152.684299188514
2011-04-18 17:22:00  327.579073188405
2011-04-18 17:52:00  156.826945856169


Please place output files in challenge2/analysis/YOURNAME/.

  1. visualization of the data-set as values per 30mins
  2. visualization of the data-set as values per day
  3. visualization of the data-set as clusters

Challenge 3 - Web Application

Please create a web application to show house energy production.

  1. Insert csv files into SQL database. (MySQL, postgreSQL, etc..)
  2. Load data from DB and show it on the web with a web framework. (Rails preferred)
  3. Show 1 or 2 types of charts of the data. (no more than 2 types)
  4. (Option) Deploy it somewhere. (AWS, Heroku, your own server, etc...)

We will review basic programming skill, data modelling and what to show. We will not review your web design skill.


Input dataset files in the challenge3/data/ directory contain data in the following format:

$ ls data/
dataset_50.csv  house_data.csv

$ head data/house_data.csv | column -s, -t
ID  Firstname  Lastname  City       num_of_people  has_child
1   Carolyn    Flores    London     2              Yes
2   Jennifer   Martinez  Cambridge  3              No
3   Larry      Robinson  London     4              Yes
4   Paul       Wright    Oxford     3              No
5   Frances    Ramirez   London     3              Yes
6   Pamela     Lee       Oxford     3              Yes
7   Patricia   Taylor    London     3              Yes
8   Denise     Lewis     Oxford     4              Yes
9   Kelly      Clark     Cambridge  4              No

(Names are by Random Name Generator )

dataset_50.csv is almost same to Challenge1's Input. It is smaller and its ID starts with 1 rather than 0. Please refer to Challenge1. house_data.csv is household data related to dataset_50.csv. The first line gives the format name. ID column values in this file are same to House column values in dataset_50.csv. City column includes 'London', 'Cambridge' and 'Oxford'. has_child column has only 'Yes' or 'No'.


  1. Fork this repository.
  2. Place all source code in challenge3/webapp/YOURNAME/.
  3. Create a pull request.
  4. (Option) Write deployed url in pull request comment.

You can refer to sample implementation in challenge3/webapp/sample/, but please bear in mind that it is a rough implementation and may be broken in some places.

Challenge 4 - WEB-API Server

Please create a web api server to calculate electricity charges.

  1. see TEPCO's explanation of electricity charges.
  1. you have to calculate Energy Charge of "Meter-Rate Lighting B" and "Yoru Toku Plan", and write WEB-API server with your favorite web framework ( Ruby on Rails preferred ) Energy Charge grows when the energy consumption ( kWh ) is bigger.

  2. deploy it to somewhere ( AWS, heroku, your own server, etc...)

We will review basic programming skill, API design and performance.


The input dataset files in challenge4/data/ contain data in the following format:

  $ ls data/
sample-consumption.json plans.json

  $ cat data/sample-consumption.json
  [ 0.2, 0.3, 0.2, ... ], # 24 values for 1st day, 1 am, 2 am .. 12 am, 1 pm ..  12 pm
  [ 0.2, 0.3, 0.2, ... ], # 24 values for 2nd day
  [ 0.2, 0.3, 0.2, ... ]  # 24 values for 31st day

sample-consumptions.json is a JSON array of arrays of float values. Each float value is a energy consumption(kWh). First value of a day is a consumption from 0 am to 1 am.

  $ cat data/plans.json
  "Meter-Rate Lighting B": {
    "Day time": [
      [ null, 120, 19.43],
      [ 120, 300, 25.91],
      [ 300, null, 29.93]
    "Night time": null,
    "Night time range": null

  "Yoru Toku Plan": {
    "Day time": [
      [ null, 90, 24.03],
      [ 90, 230, 32.03],
      [ 230, null, 37.00]
    "Night time": [
      [ null, null, 12.48]
    "Night time range":
      [ true, true, true, true,
        true, false, false, false,
        false, false, false, false,
        false, false, false, false,
        false, false, false, false,
        false, true, true, true ]

"Day time" and "Night time" values are array [ from kWh, to kWh, unit price tax included ]

[ null, 120, 19.43 ] :
means the unit price is ¥19.43 per kilo watt hour upto initial 120 kWh.
[ 300, null, 29.93 ] :
means the unit price is ¥29.93 when the energy consumption is larger than 300 kWh.

When "Night time" attribute is null, the plan has only day time. "Night time range" is 24 boolean values which represent 24 hours night time and day time.

(Yoru Toku Plan offers discount rate in night time.)

Output of the API is just one float number of Energy Charge with tax.


  1. Please set ALL source codes in challenge4/webapp/YOURNAME/
  2. Write deployed URL in Pull Request Comment.