Please use Python>=3.5.
To start the challenge interactively(recommended):
pip install -r requirements.txt
jupyter notebook
and open the challenge notebook.
If you prefer to just work on a script:
pip install pandas
python challenge.py
Your solution will be evaluated against the following criteria:
- pandas/numpy knowledge
- complexity (the lower the better)
- efficiency (use of GIL releasing / cythonized functions, memory use, etc.)
- ease of parallelization
- readability (PEP8 Standard)
- documentation
Due to high submission count with great differences in solution quality we evaluate all challenges first automatically.
- 30 points for correctness
- 30 points for speed rel to 60 sec.
- 20 points for memory rel to 1GB
- 10 points for doc
- 10 points for flake8 compliance
Please don't generate testdata in your submission. Your solution will be evaluated automatically and checked for correctness using an automated testuite on real data.
The SUM_INTERVAL variable contains relative time intervals in days.
So the first means 7 days back until today (asof time writing this 2016-02-01).
The generated files contain 1000 distinct users which made visits over the timespan of one month.
Each visit has a value for feature a, as awell as feature_b assigned to it.
For each user calculate the sum of it's respective features for each time interval. The final output should be a dataframe or numpy matrix containing one row per user, an id column an the feature columns (N_users, 1 + N_intervals*N_features)
The feature columns are expected to be in descending order (in terms of days included, so 30 days feature A as the first column and 30 days feature B is second and so on). The resulting matrix should be ordered by the user ids.
When the relative interval (-7, 0) is given with the date 2017-02-01 the semantics
are to include features values within following time interval:
[2016-01-25 00:00:00; 2016-02-01 00:00:00)
You are encouraged to use the pandas library for this task but it is not required.
Please zip your solution including all files (do not include the generated data directory!) and send to us with the following naming schema:
cc_<name>_<last_name>.zip
Send each code challenge as a separate zip file!
Solutions that don't follow this convention can not be evaluated.