Cryptoledger is a tool to keep track of cryptocurrency transactions
- Imports *.csv transaction lists exported from exchanges (bittrex and gdax supported) and consolidates them in a standardize *.csv file without duplicates.
- Calculate a portfolio on any given dates using the transaction list
- Estimate its value in specified currency using cryptocompare API
- Plot a portfolio pie chart, daily valuation and daily return for a given period
- Create a corresponding PDF report formatted with a Latex template
- A working Latex distribution with pdflatex must be installed to compile the final pdf report
- This program was tested on linux, using python 3.6.3. Other operating systems might require changes
Install from github
pip3 install git+https://github.com/rfortiz/cryptoledger
To import in other python scripts
from cryptoledger import cryptoledger
python3 -m cryptoledger.cryptoledger ledger_db --export
Loads a ledger database (*.csv) and export a pdf report in the same folder
All imported transactions are consolidate in a user specified *.csv file. A new file is created if it doesn't exist. Regardless of the source used for import, all transactions are saved in a csv file with the following format:
| UID | Base currency | Quote currency | Action | Quantity (#base) | Rate (Quote currency) | Amount (Quote currency) | Commission (Quote currency) | Timestamp | Exchange |
|---|---|---|---|---|---|---|---|---|---|
| transaction_id | ETH | BTC | BTC | BUY | 3 | 0.03133112 | 0.09399336 | 0.000783278 | 1512651720 |
- UID transaction identification number used on exchange or blockchain for transfers (DEPOSIT|WITHDRAW)
- Base currency symbol of the base currency
- Quote currency symbol of the quote currency (equal to Base currency for transfers)
- Action one of four actions: BUY|SELL|DEPOSIT|WITHDRAW
- Quantity quantity of the base currency
- Rate exchange rate (equal 1.0 for transfers)
- Amount = Quantity * Rate, does not include commission
- Commission fees from exchange or on chain transaction
- Timestamp unix timestamp of the transaction closing date
- Exchange name of the exchange in small caps
python3 -m cryptoledger.cryptoledger ledger_db --add csvs_folder_path
imports all .csv files in csvs_folder_path folder. Filnames must follow: ** exchangeName_.csv **, where exchange name is 'bittrex', 'gdax' or 'native' to import another database following the format above.
to get *.csv files from exchanges:
- bittrex: Orders → scroll down to 'Completed' → Load all (do not click 'csv')
- **gdax: ** Fills → select currency pair → Download Receipt/Statement
python3 -m cryptoledger.cryptoledger ledger_db --manual
Transactions can be added manually from command prompt with the argument -m. This can be used to add withdrawals or deposits that are not exported by the exchanges. Alternatively, a csv file named native_*.csv following the format above can be manually generated and imported using the -add argument
python3 -m cryptoledger.cryptoledger ledger_db --export
generates a pdf report in EUR from the date of the first transaction to the current date. The valuation currency can be changed with --currency:
python3 -m cryptoledger.cryptoledger ledger_db --export --currency USD
and dates can be selected with --start_date/--end_date
python3 -m cryptoledger.cryptoledger ledger_db --export --start_date 2017-12-01 --end_date 2017-12-31
-start/end dates -currency -files generated in same folder as ledger
This report example can be reproduced with the dummy *.csv files provided in ./bin/csv_import. Assuming csv_import folder is in the working path:
cd [working path where generated files will be stored]
python3 -m cryptoledger.cryptoledger ledger.csv --add ./csv_import --export
- Pandas - Datastructures used to handle the protfolio
- Matplotlib - To create the plots
- Cryptocompare API - historical cryptocurrency data
- Jinja2 - Templating engine used to parse Latex file and insert python variables
- LaTeX - Document preparation system used to creat PDFs
This project was undertaken to learn python programming and several aspects could be improved:
- The ledger (transactions list) is handled by a custom class using python built-in functions. Using Pandas dataframe would be simpler and more efficient.
- More error handling (user inputs, invalid transactions, api offline, etc.)
- Test other operating systems, python version, etc.
- Avoid recalculating the portfolio from ledger every time (not a problem for small database, time consuming currency conversion using cryptocompare API is saved and reused until new transactions are added)
- Daily portfolio DataFrame contains lots of duplicates and could be replaced by Pandas' sparse data structures
- Tracking unmatched transactions
This project is licensed under the MIT License - see the LICENSE.md file for details