Skip to content

Commit

Permalink
Release v1.0 of the schema (#26)
Browse files Browse the repository at this point in the history
  • Loading branch information
odow authored Oct 28, 2023
1 parent 96b6671 commit 120e039
Show file tree
Hide file tree
Showing 12 changed files with 339 additions and 434 deletions.
1 change: 1 addition & 0 deletions .gitignore
Original file line number Diff line number Diff line change
Expand Up @@ -2,3 +2,4 @@ __pycache__
*.pyc
.DS_Store
sol_*.json
examples/lang-julia/Manifest.toml
14 changes: 5 additions & 9 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -7,10 +7,7 @@ optimization problems called _StochOptFormat_, with the file extension
For convenience, we abbreviate StochOptFormat to _SOF_.

StochOptFormat is defined by the [JSON schema](http://JSON-schema.org)
[`https://odow.github.io/StochOptFormat/versions/sof-0.3.schema.json`](https://odow.github.io/StochOptFormat/versions/sof-0.3.schema.json).

_Note: StochOptFormat is in development. If you have suggestions or comments,
please [open an issue](https://github.com/odow/StochOptFormat/issues/new)._
[`https://odow.github.io/StochOptFormat/versions/sof-1.schema.json`](https://odow.github.io/StochOptFormat/versions/sof-1.schema.json).

**Authors**

Expand Down Expand Up @@ -92,9 +89,8 @@ When developing StochOptFormat, we set out to create:
Equally as important as the things that we set out to do, are the things that we
did _not_ set out to do:

- We did not try to incorporate chance constraints
- We did not try to incorporate continuous random variables
- We did not try to incorporate decision-hazard nodes.
- We did not try to incorporate chance constraints.
- We did not try to incorporate continuous random variables.

Finally, StochOptFormat is not an algebraic modeling language for stochastic
programming. Instead, it is an instance format [5].
Expand Down Expand Up @@ -249,7 +245,7 @@ Encoded in StochOptFormat, the newsvendor problem becomes:
"name": "newsvendor",
"date": "2023-05-02",
"description": "A StochOptFormat implementation of the classical two-stage newsvendor problem.",
"version": {"major": 0, "minor": 3},
"version": {"major": 1, "minor": 0},
"root": {
"state_variables": {"x": 0.0},
"successors": {"first_stage": 1.0}
Expand Down Expand Up @@ -548,7 +544,7 @@ resulting policy, such as expected objective values, and various quantiles.
trivially extends to infinite horizon problems and problems with a stochastic
process that is not stagewise independent.

- Q: MathOptFormat is too complicated. Why can't we use LP or MPS files?
- Q: Why MathOptFormat instead of LP or MPS files?

A: MathOptFormat can be read and writen by most programming languages. In
addition, it is very general and easy to extend. Please read Section 2 of [2]
Expand Down
Loading

0 comments on commit 120e039

Please sign in to comment.