README.md

Nectar Data Challenge

You have 2 data sets - one data set is collected from the sensors (used.csv) and another one is collected from the point of sale system (sold.csv).

I would want you to correlate the 2 data sets using the following rules:

• you should match rows based on the product_id (this is the product that is being tracked by a given sensor or sold), time (this is the time when the products was used or sold) and used_ml/sold_ml (the amount that was used/sold) using the following rules:

• product_id should be matched 1:1 btw used and sold

• time should be within 20 minutes btw sold and used

• pick the closest used_ml vs sold_ml within the time window to get a 1:1 match

• There is some amount of nuance to the matching - don't worry too much about it, as long as it is an approximation of the rules.

There will be rows that won't match - you can simply ignore those.

Plot the distribution of the absolute difference between the used_ml and sold_ml, aggregated across all products and times.

Feel free to use any tools/programming languages you are most comfortable with. Also, feel free to ask for any clarifications you need.

Happy hacking!