Run a model and inspect its result

This guide continues Get started with the same source revision, decay.eqi, run.py and uv environment. If those files are not ready, follow that page’s installation and downloads first. No additional dependency is needed.

Follow the script

The maintained script on Get started has four responsibilities:

1. eqiora.compile(path=...) reads the equations and returns a Model.
2. eqiora.resolve(...) chooses the time method independently of the model.
3. eqiora.State.initial(...) and eqiora.run(...) start from the declared initial value and request observations.
4. result.series(...) selects the time and value arrays for the model’s x field.

See How Eqiora fits together for the larger picture.

Read the output before plotting

In the folder containing the two files, run the same command:

bash

uv run --no-project --python .venv/bin/python python run.py

The script prints three observations. Times are in seconds; x is dimensionless. With rate 1, the last value is about 0.367879 at one second. That agrees with $x(t)=\exp(-t)$, rather than merely looking plausible on a plot. The script prints to the terminal and does not write a result file.

Change one thing

Use Get started’s rate edit to double the rate while leaving the script and numerical method unchanged. Predict $x(1)=\exp(-2)\approx0.135335$, then rerun the same command.

Restore the rate to 1 before following the ODE lesson and exercises. Its derivation explains why the result changes and what the model assumes.

Diagnose a failed run

Check the installation and source-diagnostic checklist first. A missing file, unresolved symbol and inconsistent unit are different problems; changing a solver setting does not repair the latter two.

Move to a spatial problem

Choose a complete Gallery case and keep its model, geometry, boundary names and numerical setup together. The current-source modeling guide gives the detailed Python operations from the same source revision.