Finac - financial accounting for humans

Finac is a library and function set for Jupyter/ipython, which provides the double-entry bookkeeping database.

Finac is simple, open and free. It can work with SQLite or any database supported by SQLAlchemy (tested: SQLite, MySQL, PostgreSQL).

Finac can be used either in the interactive mode with Jupyter, Spyder-IDE, ipython or other similar environment or Finac library can be embedded into 3rd party projects. The library can be used in accounting applications and is useful for fin-tech services.

Finac supports multiple currencies, simple transactions, double-entry bookkeeping transactions, watches overdrafts, balance limits and has got many useful features, which make accounting simple and fun.


pip3 install finac




from 0.4.10

ALTER TABLE transact ADD service bool;
UPDATE transact SET service=true WHERE d_created<'1970-01-03';
ALTER TABLE transact ADD FOREIGN KEY(chain_transact_id)
  REFERENCES transact(id) ON DELETE SET null;

from 0.3.x

Starting from 0.4, Finac uses DateTime columns for:

  • asset_rate.d
  • transact.d
  • transact.d_created
  • transact.deleted

Depending to the database type, it’s REQUIRED to convert these columns to either DATETIME (SQLite, for MySQL DATETIME(6) recommended) or TIMESTAMPTZ (PostgreSQL, with timezone).

How to use in interactive mode

Finac database contains 3 entity types:

  • asset currency, ISIN, stock code etc., currencies “USD” and “EUR” are created automatically. Finac does not separate assets into currencies, property and other. This allows creating applications for various areas using the single library.
  • account bank account, counterparty account, tax account, special account etc. Everything is accounts :)
  • transaction movements from (credit) / to (debit) and between accounts

Assets have got rates - the value of one asset, relative to other.

Transactions can be simple (no counterparty) or classic double-entry bookkeeping (between debit and credit account).

import finac as f
# init finac,
# create a couple of accounts
f.account_create('acc1', 'USD')
f.account_create('acc2', 'USD')
f.account_create('depo', 'USD', 'saving')
# import initial balance with a simple transaction'acc1', 10000, tag='import')
# move some assets to other accounts'acc2', ct='acc1', amount=2000)'depo', ct='acc1', amount=3000)
# display statement for acc1'acc1')
id     amount  cparty  tag     note  created              completed
7   10 000.00          import        2019-10-26 03:04:02  2019-10-26 03:04:02
8   -2 000.00  ACC2                  2019-10-26 03:04:02  2019-10-26 03:04:02
9   -3 000.00  DEPO                  2019-10-26 03:04:02  2019-10-26 03:04:02
Debit turnover: 10 000.00, credit turnover: 5 000.00

Net profit/loss: 5 000.00 USD
# display summary for all accounts
account  type      asset     balance  balance USD
ACC1     current    USD     5 000.00     5 000.00
ACC2     current    USD     2 000.00     2 000.00
DEPO     saving     USD     3 000.00     3 000.00
Total: 10 000.00 USD
# display summary only for current accounts'current')
account  type     asset     balance   balance USD
ACC1     current    USD     5 000.00     5 000.00
ACC2     current    USD     2 000.00     2 000.00
Total: 7 000.00 USD
# display assets pie chart, (wrapper for matplotlib.pyplot, requires Jupyter,
# Spyder-IDE or a similar interactive environment)

Note: when addressing currencies and accounts both in interactive and API mode, account and asset codes should be used as object identifiers. All codes are case-insensitive.

Inside database Finac uses numeric IDs to connect objects, so the codes can be changed without any problems.

Special features

Lazy exchange

Finac can automatically move assets between accounts having different currencies if exchange rate is set or specified in the transaction details:

# create EUR account
f.account_create('acc5', 'eur')
# set exchange rate (in real life you would probably use cron job)
f.asset_set_rate('eur/usd', value=1.1)'acc5', ct='acc1', amount=100)

hoorah, account acc5 have got 100 EUR! And exchange rate was 1.1. Check it:

id     amount  cparty  tag     note  created              completed
14    -110.00                        2019-10-26 03:15:41  2019-10-26 03:15:41
id  amount  cparty  tag  note  created              completed
15  100.00                     2019-10-26 03:15:41  2019-10-26 03:15:41
Debit turnover: 100.00, credit turnover: 0.00

Net profit/loss: 100.00 EUR

As shown, there is no a counterparty account in the lazy exchange. This feature is useful for personal accounting and special applications, but for professional accounting, create counterparty exchange accounts should be created and buy-sell transactions should be performed between them.


Targets is a feature I have created Finac for. Consider there are account balances in a bank and in the accounting. They differ in some amount and this need to be recorded in the accounting with a single transaction.

But the problem is: there is a lot of transactions which should be sum up. Or the difference between bank balance and accounting must be calculated manually. Pretty common, eh? Don’t do this, Finac has got targets.

Specifying targets instead of amount asks Finac to calculate transaction amount by itself.

After the previous operation, there is 4,890.00 USD on “acc1” and consider all except $1000 should be moved to “acc2”. Let us do it:

>>>'acc2', ct='acc1', target_ct=1000)
id     amount  cparty  tag     note  created              completed
16  -3 890.00  ACC2                  2019-10-26 03:25:56  2019-10-26 03:25:56
Debit turnover: 10 000.00, credit turnover: 9 000.00

Net profit/loss: 1 000.00 USD

The transaction amount is automatically calculated. Lazy people are happy :)

If the debit account balance target should be specified, target_dt function argument can be used. Note: calculated transaction amount must be always greater than zero (if credit account target higher than its current balance is specified, ValueError is raised)

For simple transactions (…))), use target=.

Transaction templates

Example: there is a repeating payment orders in a bank, which pay office utility bills every 5th day of month, plus automatically move $100 to a saving account. To fill this into accounting, YAML transaction template can be used:

  - account: acc1
    amount: 200
    tag: electricity
    note: energy company deposit
  - account: acc1
    amount: 800
    tag: rent
    note: office rent
  - dt: depo
    ct: acc1
    amount: 200
    tag: savings
    note: rainy day savings

then create a cron job which calls f.transaction_apply(“/path/to/file.yml”) and that is it.

Actually, transaction templates are useful for any repeating operations. The same arguments, as for the core functions, can be specified.

Number formatting

Finac does not use system locale. If amounts and targets are inputted as strings, they can be specified in any format and Finac tries converting strings into float numeric automatically. The following values for amounts and targets are valid and are automatically parsed:

  • 1 000,00 = 1000.0
  • 1,000.00 = 1000.0
  • 1.000,00 = 1000.0
  • 1,000.00 = 1000.0
  • 10,0 = 10.0
  • 10.0 = 10.0

Passive accounts

If account is passive, its assets are decremented from totals. To create passive account, passive argument must be used:

f.account_create('passive1', 'usd', passive=True)

Accounts of types “tax”, “supplier” and “finagent” are passive by default.

Data multiplier

Depending on data, it may be useful to store numeric values in the database as integers instead of floats. Finac library has got a built-in data multiplier feature. To enable it, set multiplier=N in finac.init() method, e.g. multiplier=1000. This makes Finac to store integers into tables and use the max precision of 3 digits after comma.

Note: table fields must be manually converted to numeric/integer types. In production databases the field values must be also manually multiplied.

Full list of tables and fields, required to be converted, is available in the dict finac.core.multiply_fields.

Note: the multiplier can be used only with integer and numeric(X) field types, as core conversion functions always return rounded value.

How to embed Finac library into own project

See Finac documentation for core function API details.

Client-server mode and HTTP API

See Finac documentation for server mode and HTTP API details.

Enterprise server and support

Want to integrate Finac into an own enterprise app or service? Need a support? Check Finac Enterprise Server.