From 7249dc87802af5a4895d7131b983cba3e5bdc233 Mon Sep 17 00:00:00 2001 From: "Riet, Adriaan Anthony" Date: Wed, 3 Jan 2024 10:58:26 -0700 Subject: [PATCH 1/3] Update ParallelSubsetSimulation docs reference #26468 --- .../content/source/samplers/ParallelSubsetSimulation.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md index 696c5fa07619..a0b5d33379ac 100644 --- a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md +++ b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md @@ -105,6 +105,14 @@ chains, the `AdaptiveMonteCarloDecision` class is used for decision-making. Please refer to [AdaptiveMonteCarloDecision](AdaptiveMonteCarloDecision.md) for more details. +If the parameter `use_absolute_value = true`, the absolute value of `output_reporter` will be maximized. +As such, a least-squares fit optimization can be achieved by passing the negative of the sum-squared error +into the value of `output_reporter` with `use_absolute_value = true`. Without `use_absolute_value = true` +the value of `output_reporter` will be maximized after its absolute value is taken. + +As the PSS algorithm is iterative, an `Executioner` block with type `Transient` is required for proper +operation. + ## Example Input Syntax The input file for using the PSS algorithm is somewhat similar to the other sampler From 54d1c7671ff7cf8615539ba99a5aee638555f571 Mon Sep 17 00:00:00 2001 From: "Riet, Adriaan Anthony" Date: Wed, 3 Jan 2024 12:50:09 -0700 Subject: [PATCH 2/3] Adjusted syntax for all referenced parameters in the input file --- .../samplers/ParallelSubsetSimulation.md | 29 ++++++++++--------- 1 file changed, 16 insertions(+), 13 deletions(-) diff --git a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md index a0b5d33379ac..0b7e2e84bdd0 100644 --- a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md +++ b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md @@ -105,13 +105,13 @@ chains, the `AdaptiveMonteCarloDecision` class is used for decision-making. Please refer to [AdaptiveMonteCarloDecision](AdaptiveMonteCarloDecision.md) for more details. -If the parameter `use_absolute_value = true`, the absolute value of `output_reporter` will be maximized. +If the parameter [!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) is `True`, the absolute + value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) will be maximized. As such, a least-squares fit optimization can be achieved by passing the negative of the sum-squared error -into the value of `output_reporter` with `use_absolute_value = true`. Without `use_absolute_value = true` -the value of `output_reporter` will be maximized after its absolute value is taken. - -As the PSS algorithm is iterative, an `Executioner` block with type `Transient` is required for proper -operation. +into the value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) with + `[!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) = true`. Without +`[!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) = true` +the value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) will be maximized after its absolute value is taken. ## Example Input Syntax @@ -120,14 +120,17 @@ The input file for using the PSS algorithm is somewhat similar to the other samp !listing modules/stochastic_tools/test/tests/samplers/ParallelSubsetSimulation/pss.i block=Samplers -where, `num_samplessub` is the number of samples per subset and `num_subsets` is the number of subsets. -`inputs_reporter` and `output_reporter` are the reporter values which transfer -information between the `ParallelSubsetSimulation` sampler and the `AdaptiveMonteCarloDecision` reporter. There is an optional input parameter `subset_probability` which has been defaulted to +where, [!param](/Samplers/ParallelSubsetSimulation/num_samplessub) is the number of samples per subset and +[!param](/Samplers/ParallelSubsetSimulation/num_subsets) is the number of subsets. +[!param](/Samplers/ParallelSubsetSimulation/inputs_reporter) and [!param](/Samplers/ParallelSubsetSimulation/output_reporter) +are the reporter values which transfer +information between the `ParallelSubsetSimulation` sampler and the `AdaptiveMonteCarloDecision` reporter. There is an optional + input parameter [!param](/Samplers/ParallelSubsetSimulation/subset_probability) which has been defaulted to `0.1`, meaning that there are $1/0.1$ samples per Markov chain. This can, however, be -changed as per the user preference. `num_parallel_chains` is also an optional parameter +changed as per the user preference. [!param](/Samplers/ParallelSubsetSimulation/num_parallel_chains) is also an optional parameter that explicitly specifies the number of Markov chains to be run in parallel per subset. -If `num_parallel_chains` is not specified, the number of parallel Markov chains per subset -will be equal to the number of processors. Besides, `use_absolute_value` can be set to true +If [!param](/Samplers/ParallelSubsetSimulation/num_parallel_chains) is not specified, the number of parallel Markov chains per subset +will be equal to the number of processors. Besides, [!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) can be set to true when failure is defined as a non-exceedance rather than an exceedance. Second, the `Reporters` block is presented below with the `AdaptiveMonteCarloDecision` reporter: @@ -141,7 +144,7 @@ Third, the `Executioner` block is presented below: !listing modules/stochastic_tools/test/tests/samplers/ParallelSubsetSimulation/pss.i block=Executioner where it is noticed that unlike some other sampler classes, the `type` is transient. -The number of time steps is automatically determined based on `num_samplessub`, `num_subsets` and the number of processors used. +The number of time steps is automatically determined based on [!param](/Samplers/ParallelSubsetSimulation/num_samplessub), [!param](/Samplers/ParallelSubsetSimulation/num_subsets) and the number of processors used. ## Output format From a87b694e8fedf658720e03ff7f40932a280c2aec Mon Sep 17 00:00:00 2001 From: "Riet, Adriaan Anthony" Date: Wed, 3 Jan 2024 13:40:40 -0700 Subject: [PATCH 3/3] Fixed formatting error when adjusting parameter referencing --- .../doc/content/source/samplers/ParallelSubsetSimulation.md | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md index 0b7e2e84bdd0..258162afa388 100644 --- a/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md +++ b/modules/stochastic_tools/doc/content/source/samplers/ParallelSubsetSimulation.md @@ -109,8 +109,7 @@ If the parameter [!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) will be maximized. As such, a least-squares fit optimization can be achieved by passing the negative of the sum-squared error into the value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) with - `[!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) = true`. Without -`[!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) = true` + [!param](/Samplers/ParallelSubsetSimulation/use_absolute_value) set to `True`. Otherwise the value of [!param](/Samplers/ParallelSubsetSimulation/output_reporter) will be maximized after its absolute value is taken. ## Example Input Syntax