General rules

Questions

Split

Problem

Write a function called split which will take two iterables as input - One called items and another called ratios. It should return an iterable of iterables which contains random elements from items in the proportions as indicated in ratios. You should round up or down to make sure all elements are there in the return value.

e.g. The following invocation will split [1,2,3,4,5,6,7,8,9,10] into 3 lists. The first one will have 5 random elements (50%), the next will have 4 (40%) and the last will have 1 (10%).

  split([1,2,3,4,5,6,7,8,9,10], [0.5, 0.4, 0.1])

will return something like

   [[4,2,3,9,10], [1,5,6,8], [7]]

If instead, we called it like this

    split([1,2,3,4,5,6,7,8,9,10], [0.25, 0.5, 0.25])

will return something like

      [[2,3,5], [1,4,6,7,9], [8,10]]

Notice that the 0.25 entries have different values to accommodate all the elements.

Rules

1. Please don't use any 3rd party libraries to do this. No pandas, numpy etc.
2. Make sure your code is well structured and easy to read. Good variable names, proper functions etc will go a long way.
3. Include unit tests for your submission

Crawler

Problem

Write a program called crawler.py to scrape the website https://www.gettyimages.in/photos/aamir-khan-actor and download all 60 images of the actor.

• Each image will have a URL like this https://media.gettyimages.com/photos/indian-actor-amir-khan-1988-picture-id499342585?k=6&m=499342585&s=612x612&w=0&h=4jP2Zn9qP_R5kN8NpOwTTCDBO2SEVeqfZDcF0VebaKw=. This what you need to download. They will be linked to a larger image like https://media.gettyimages.com/photos/indian-actor-amir-khan-1988-picture-id499342585?s=2048x2048. You don't need to download that. Only the smaller former one.
• After you download an image, save it in a directory called images. The filename has to be created from the input URL using the pathname. E.g. if the URL from which you downloaded is https://media.gettyimages.com/photos/indian-actor-amir-khan-1988-picture-id499342585?k=6&m=499342585&s=612x612&w=0&h=4jP2Zn9qP_R5kN8NpOwTTCDBO2SEVeqfZDcF0VebaKw= the filename will be called indian-actor-amir-khan-1988-picture-id499342585.jpg. This is the pathname of the URL without the /photos prefix.
• At the end of the operation, you will have a directory called images that contains 60 images of the actor with the correct names and checksums.

Click on the following image to see an ascii screencast of how this will work.

Rules

1. Use whatever libraries you need for this. But please provide a
   proper `requirements.txt` and a `setup.py` file. 

Statistics

Problem

The attached file called unemployment-rate.csv contains a list of employment ratios for various countries for various years. Each entry looks like this

    Belgium,BEL,2010,8.28999996185303

Where Belgium is the country, BEL is the symbol for the country. 2018 is the year and 8.28999996185303 is the unemployment ratio.

Write a command line program called statistics.py which should accept an input file and the following command line parameters

    --country COUNTRY  Country to perform operation for
    -o {avg,min,max}   Operation to perform on dates. (Default avg)
    --from FROM_       Starting year (inclusive)
    --to TO            Ending year (inclusive)

It should then parse the input file and filter the output as per the provided parameters and display.

The --country parameter is compulsory and will narrow the statistics to only entries for that country. It accepts country names (not symbols)

The -o option will take the operation to perform. It can be either min for minimum, max for maximum or avg for average. It will perform this operation on the unemployment rates for that country. This is not compulsory. The default (when nothing is provided is avg).

--from and --to are optional and used to restrict the operations to a date range. e.g. if you run it with --from=1999, it will only consider the lines with the year greater than or equal to 1999. Similarly, --to=2000 will only consider lines less with the year than or equal to 2000.

The output will simply print the requested statistic.

Click on the following image to see an ascii screencast of how this will work.

Rules

1. Don't use anything except the standard library to do this. No pandas, numpy etc.
2. Make sure that the code is modular and has separate functions to do various tasks
3. Unit tests are compulsory. Please include them.
4. Use an argument parsing library to handle the command line parsing

Evaluation criteria

1. Modular code. We expect you to write proper functions for each task and then put them together rather than one single program
2. We expect a proper requirements.txt file and a setup.py file so that the program can be installed and used
3. (optional) If you include a test suite, you will be awarded extra points
4. Good code style - Proper variable names, no misspellings, adherence to Python common coding standards will all be considered.
5. Please version control your development and send us a git repository with your code. It's important that we see the process of development so please don't just upload everything in a single shot.