Skip to content

Files

Latest commit

ashish-kambojashish-kamboj
Add files via upload
Mar 14, 2021
edadae9 · Mar 14, 2021

History

History

Amaze Payment Solution EDA

README.md

Problem Statement:

Amaze Payment Solution (APS) has launched its product in Feb'16. Using this product, users can transact online on various merchants with the credit amount which APS provides. APS has bimonthly credit cycle and dates being 1st and 16th of every month. On these dates bill will be generated for the user of the transaction amount they have transacted in the cycle. Users can pay APS back after the bill generation and continue using the service. Although, users can also do 'onetime_settlements', which means to pay APS before the bill generation of the cycle. If user doesn't pays APS after the bill generation, the bill remains in pending state and user cannot transact until he/she clears his/her dues. APS has data from the launch. Now, they want to predict the repayment behaviour of the users who transacted in the latest cycle (id = 22).

Goal:

You are expected to do EDA on the data and generate a report on variables which are important for user repayments. Also, you have to share us findings, insights and recommendations on the basis of EDA.

Datasets:

  • credit_data
  • cycles
  • transactions_data
  • users_data
  • settlements_data
  • failure_events_data

  • credit_data: This dataset has user's credit limit data at cycle level. Users's credit limit gets updated depending upon his transactional and settlements behavior. This dataset gives information of user's credit for a cycle.

    • user_id: Unique indentifier of a user
    • cycle_id: Unique indentifier of a cycle
    • global_credit_limit: Credit limit of a user

  • cycles:

    • cycle_id: Unique indentifier of a cycle
    • start_date: Timestamp when cycle started
    • end_date: Timestamp when cycle ended

  • transactions_data:

    • transaction_id: Unique indentifier of a transaction
    • user_id: Unique indentifier of a user
    • merchant_id: Unique indentifier of a merchant trasnsacting with
    • transaction_amount: amount of transaction
    • created_at: transaction timestamp

  • users_data:

    • user_id: Unique indentifier of a user
    • referrer: User approval source
    • email: email id of user
    • name: user name
    • city_id: city from which user transacts

  • settlements_data: This data is has user's repayments related variables

    • settlement_id: Unique indentifier of a settlement
    • user_id: Unique indentifier of a user
    • cycle_id: Unique indentifier of a cycle
    • settlement_amount: repayment amount paid by user
    • settlement_status: type of settlement (bill_pending: if bill is due
    • bill_settled: if bill is settled
    • onetime_settlement: paid before bill generation)
    • days_delayed: number of days repayment delayed
    • settlment_created_at: bill generation timestamp or onetime_settlement creation timestamp
    • settlement_updated_at: bill paid timestamp or onetime_settlement paid timestamp

  • failure_events_data: This data has errors happened while transaction. Due to these errors transaction was interrupted

    • failure_event_id: Unique indentifier of failure event
    • transaction_id: Unique identifier of transaction
    • user_id: Unique indentifier of a user
    • error_type: type of error occured while transacting
    • amount_in_paise: amount of transaction in while error occured
    • created_at: timestamp when error occured