🏛️ Accounting

Aim is to propose a simple algorithm that read, parse, store, process and display financial data for individuals 💸.

🎉 Examples

Example in old version (Click to expand)

🗓️ Dates

🚀 Version 2 Started

Project pitched and started the 11th december 2020

🎯 Release date

• First expected release the friday 15th december 2020
• Second expected release the 10th january 2021

🔌 Dependencies

1. 🖥️ Set your Virtual Environment:

   # Download venv librairy
   apt-get install python3-venv -y
   # Create your venv
   python3 -m venv venv
   # Activate your venv
   . venv/bin/activate

   For more information, go to Python Virtual Environment Official Documentation.

2. 📦 Install the project dependencies:

   apt install python3-pip
   pip install PTable
   pip install Flask
   
   pip install PyPDF2

⚡ Quick start

1. Database configuration:

   	mkdir -p back/database/ && touch back/database/accounting.db
   	mkdir -p data/sg/ 
   	# create the input directory for file integration

2. To start the backend server, simply run bellow commands:

   	export FLASK_APP=controller.py
   	export FLASK_ENV=development
   	flask run
   	# * Running on http://127.0.0.1:5000/

🎨 Architecture

Version 2

Version 1 (Click to expand)

📋 Tasks

1. Database

   • Create schema
   • Try to use some DB framework
   • Fix the duplicate issue

2. Payments

   • Create account
   • Fake few payments
   • Compute few metrics

3. REST API

   • Expose data

     • Flask
     • GraphQL ~> not very effective

   • Design API (Postman ~> not effective)

   • Insert payments ?

   • Expose metrics

4. Integration

   • CSV from Societe Generale
     • Get all files in directory
     • read each file, one after the other
     • Parser rows and map to RawPayments
   • Save in database
     • Generic object's management in database
     • If file is duplicate => skip
   • View all payments and duplications between files (count same lines in file and across files)
   • CSV from N26
   • 4% of the database (72/1685 payments) is corrupted by duplicated values from different files

5. Statistics

   • Low level stats
     • Total credit/debit for quantity and amount
     • Min, Max, Avg, opening, closing
   • Handle consumption periods (~daily, weekly, monthly)
     • Add periodicity: Trimester, Semester
   • Metrics column info
     • Add header info line for metrics column
     • Add footer avg line for metrics column
   • Identify FIXED income and expenses (try to rely on LABELS)
   • Define profile
   • string similarity calculation to identify recurrent transactions

6. Filter v2

   • Allow labeling (Add, remove, reset)
   • Auto labeling on token
     • Available labels: transport, rent, salary, phone, AMZ, SG, Healthcare...
   • Filter payments on account & list of labels
   • Ensure payments uniqueness
   • Group some labels together (expose groups to client ?)
   • Filter payments on account & group
   • Percentage of payments (qty & amount) labeled (or group) for a given period
   • Pie chart for example
   Filter v1 (Click to expand)

   1. Labels

      • Tag all payments

      • Define groups based on tags

      • Create groups that matches perfectly one tag, multiple tags or other groups

      • Apply metrics on groups

      • Allow enforced new tag for payment id

      • Count payments in groups

      • Count payments untagged

      • Identify overlapping tags

      • Auto labeling on rules

      • Endpoint for labeling

      • Create categories and pattern that fall in this category

        • Courses
        • Amazon

      • Can 2 categories share same payments ?

        • Need exclusive categories for global expenses pie-chart
        • and non-exclusive

7. Integrate PDFs

   • Download all PDFs available from banking platforms
   • Read Parse (PyPDF2, tabula)
   • Extract data ~> not very effective
   • Store raw payments

8. Architecture

   • Design first diagram of internal services
     • Try coding schema
   • Add the schema to README
   • Define API specification using Swagger

9. Front

   • Consumption map on a calendar just like github contribution calendar
   • Display payments curve

🕹️ More examples

💾 Database example