lifelib: Actuarial models in Python¶
An open-source library of life actuarial models written in Python. You can run the models right out of the box, customize them in any way you want, or create your own models from scratch.
Last Updates
- 6 July 2019: lifelib v0.0.12 is released. See Releases for details.
- 24 March 2019: lifelib v0.0.11 is released. See Releases for details.
- 24 March 2019: modelx v0.0.21 and spyder-modelx v0.0.9 is released. See modelx documentation for more details.
- 2 February 2019: lifelib v0.0.10 is released. See Releases for details.
- 2 December 2018: New Using Spyder plugin sub page is added in Quick Start.
- 2 December 2018: modelx v0.0.17 and spyder-modelx v0.0.6 is released. See modelx documentation for more details.
- 11 November 2018: New notebook IFRS17 Simulation (Lapse scenario) is added.
- 3 October 2018: Update Project ifrs17sim to pay out profits each period.
- 3 October 2018:
Fix miscalculation in
IntAccumCF()inprojection.
Feature highlights
- Readable formulas
- Multidimensional data structure
- Instant evaluation
- Dependency tracking (Under development)
- Reusable code
- Object oriented models
- Interface with Excel/Pandas
- Version control
- Documentation integration
Why lifelib?
- Better model integrity and extensibility
- For readable formula expressions
- For eliminating spreadsheet errors
- For better version control/model governance
What for?
- For research/educational projects
- As communication tools to convey model specifications
- Model validation / testing
- Prototyping for production models
- Pricing / Profit testing
- As corporate models
- For simulations
- As replacement for any spreadsheet models
How lifelib works
You can create an actuarial projection model as modelx Model object by:
- Creating a project folder from a lifelib project template,
- Build a modelx Model from source modules and
input data in the project, by running
buildfunction.
Once the model is built, they are available as a modelx Model object in Python console. The model is composed of Space. Spaces contain Cells and other spaces. Cells are much like cells in spreadsheets, which in turn, can store formulas and associated values.
With a lifelib model, you can:
- Get calculated values by simply accessing model elements,
- Change the model by changing input and writing formulas in Python,
- View the tree of model elements in graphical user interface,
- Output results to Pandas objects,
- Save the model, load it back again, and do much more.
Start from Quick Start page.
Got questions?
If you have troubles to shoot, questions to be answered,
post them on stackoverflow
and add lifelib tag to the posts.
If you find bugs or want to request new features,
submit issues on
lifelib development site
on github.
Or, connect with
Fumito Hamamura
on Linkedin and send a message to him (He may take a while to respond).
Links
- modelx: A Python package for building complex models of formulas and data.
- lifelib on PyPI: lifelib’s Python Package Index page.
- Development site: Github repository of lifelib.