Right now, ARMI provides the test functionality to generate a test reactor object with loadTestReactor(). By default that uses armiRun.yaml, which generates a hex-assembly-based reactor with 9 rings and 73 assemblies.
If the developer wants a smaller reactor, they can use the method reduceTestReactorRings() to reduce the number of rings in the test reactor. But that is itself quite slow, and can slow down a test.
The only silver-bullet solution to speeding up our unit tests would be to generate a smaller test reactor from the start.
I propose the solution:
We create a loadTestReactorOneBlock(), which generates a 1-assembly, 1-block reactor using the existing refOneBlockReactor.yaml.
Then, where possible, we replace loadTestReactor() with the new function to speed up our unit tests.
Right now, ARMI provides the test functionality to generate a test reactor object with
loadTestReactor()
. By default that usesarmiRun.yaml
, which generates a hex-assembly-based reactor with 9 rings and 73 assemblies.If the developer wants a smaller reactor, they can use the method
reduceTestReactorRings()
to reduce the number of rings in the test reactor. But that is itself quite slow, and can slow down a test.I propose the solution:
loadTestReactorOneBlock()
, which generates a 1-assembly, 1-block reactor using the existingrefOneBlockReactor.yaml
.loadTestReactor()
with the new function to speed up our unit tests.