Recent Posts

Sunday, August 6, 2017

sfc_models Version 0.5 Update

I am heading into what I hope is the final stretch of the "user manual" for the sfc_models package (source: https://github.com/brianr747/SFC_models). As regular readers might suspect, it ended up being longer than I planned for. Whether or not it will be available in time for the MMT conference in September is an open question. Progress this summer was slower than I expected, but that was the result of work on other projects, spending time as a baseball assistant coach (a sport I never played at an organised level!), and travelling. The package itself is essentially frozen; I have reached what will be labelled version 0.5. The next major version will be 1.0, which is all the final changes needed to support the book.

The economic functionality within the package is relatively limited; my objective was to create a solid base that makes it easy to add new types of behaviour. To this end, I attacked the areas that appeared to cause the most programming difficulty. In particular, the multi-currency support is the most complex part of the logic. Going forward, it should be possible to emulate most stock-flow consistent (SFC) models (by adding the appropriate behaviour rules), giving the package the ability to support researchers.

It was probably a mistake to write an extensive description for a package that not a whole lot of people are using, but at the same time, it is an extremely novel package (or so I hope!). It is designed to be easy to use and pick up, which implies that everything is going on under the hood. An experienced Python programmer should be able through the logic, but my guess is that most of my target audience have much less programming experience. The user guide is aimed at readers who want to understand how the package works. (The idea is that it should be easy to understand how a model is put together using the framework, but its operation would likely be a black box to non-programmers.)

With the manual out of the way, I will be attempting to drum up interest in the package. My presentation at the MMT Conference is what I will be focusing on; needless to say, I will be happy to discuss the package with people there.

There is a certain amount of background on the theory of SFC models, so it may be of interest to some of my readers who do not wish to delve into programming. However, that aspect has been much reduced as a result of the length of the more technical parts of the book. The next book in this area would be a discussion of business cycle analysis using the package; it will focus just on the theory and the model results. In my opinion, that is when the package will look more interesting, as it will be implementing more complex models.

In terms of new features, there was a syntax change introduced in version 0.5. This syntax change has the effect of breaking existing example code. I held off putting the syntax change into the production ("master") and development branches, but it has been merged into those branches. The other developments revolved around example code and multi-currency models.

(c) Brian Romanchuk 2017

No comments:

Post a Comment

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.