Donation Analytics

General Idea

I use python to solve the problem, and there is a source code called donation-analysis.py in the source file. My program will read records in incont.txt line by line. When the program read a new record, it will do:

1. check whether all fields that we are interested in are valid
2. if all fields are valid, store the record in two hash tables called transaction_record and donor_list
3. check the number of record in donor_list with key value NAME + ZIP_CODE
4. if there are more than one record in donor_list[NAME + ZIP_CODE], the donor is a repeat donor, go to step 5, otherwise, go to step 6
5. using CMTE_ID , ZIP_CODE and year as a key, call the function called printrecord() to output the requested calculations
6. repeat step 1-5 until read the end of file

Date Structure

I use two hash tables called transaction_record and donor_list to store records.

• In transaction_record, if contributions have same recipient, zip code and calender year, they will be listed together. The key of transaction_record is CMTE_ID+ZIP_CODE+year, and the corresponding value are lists of ( or maybe only one list ) [NAME,TRANSACTION_MT, month_date]
  
• In donor_list, all contributions from same donor will map to same key. The key is NAME + ZIP_CODE, and the value are lists (or maybe one list) of [CMTE_ID,TRANSACTION_MT,year,month_date]

People may wonder why we store same record two times. Although creating two hash tables which store same data may waste some space, the time complexity can reduce a lot when we find data we need and output the requested calculations.

Function

All functions are in a class called donation_analysis(), what they have to do is:

• load_percentile(): It load the percentile from percentile.txt and store it in self.percentile
• find_percentile_ind(): It find the ordinal rank through the nearest-rank method
• process_transaction_data(): load records from incont.txt and store them in two hash tables. It will called printrecord() when the new record is from a repeat donor.
• printrecord(): Given CMTE_ID,ZIP_CODE and year It uses priority queue to sort the contribution from repeat donors by the transaction month and list all output data.
• isValid(): check whether the new record has correct format

How to run the program

There is only one program called donation-analysis.py in the source file. It doesn't require additional libraries, environments, or dependencies, Users can run the program through run.sh.