OperationalLoadScenario
Economics scenarios · Implementation: mlsysim/engine/scenario_definitions/economics.py.
Construct this scenario with keyword inputs. The fields below show the types and defaults declared by this class.
Assumed operational-load shapes for N models: alerts a N, deployments N ln N, and pairwise conflicts c N^2 against one team’s team_capacity.
The shapes are a sketch, not measurements; crossing_models is the first model count whose total reaches the capacity.
Notes
Equation: The constructed load is a*n + n*ln(n) + c*n**2, compared with team_capacity on an integer grid.
Assumptions: Common abstract work units and a capacity crossing within 1..max_models.
Excludes: Empirical staffing forecasts and confidence intervals; if no crossing exists the current argmax selection returns the first grid point.
References
Derivation: The linear alert term, n*ln(n) deployment term, and quadratic conflict term are deliberately constructed workload shapes, not fitted laws. Adding them is meaningful only because the caller places them in common work units; the integer grid then locates a crossing of the supplied team capacity.
Examples
>>> from mlsysim import Scenarios
>>> s = Scenarios.OperationalLoadSketch
>>> len(s.model_counts()) == s.max_models
TrueAttributes
| Name | Description |
|---|---|
| alerts_per_model | |
| conflict_coefficient | |
| crossing_models | First grid count selected by the capacity-crossing mask; callers must ensure a crossing exists. |
| max_models | |
| team_capacity |
Methods
| Name | Description |
|---|---|
| alerts | Illustrative alert work proportional to model count; scalar or array input is accepted. |
| conflicts | Illustrative coordination work proportional to squared model count. |
| deployments | Illustrative deployment work n*ln(n); this is a chosen growth shape, not measured cost. |
| model_counts | Integer model-count grid from one through max_models. |
| total | Sum of alert, deployment, and conflict work in the scenario’s common capacity units. |
alerts
engine.scenario_types.OperationalLoadScenario.alerts(n)Illustrative alert work proportional to model count; scalar or array input is accepted.
Equation: alerts_per_model * n.
conflicts
engine.scenario_types.OperationalLoadScenario.conflicts(n)Illustrative coordination work proportional to squared model count.
Equation: conflict_coefficient * n ** 2.
deployments
engine.scenario_types.OperationalLoadScenario.deployments(n)Illustrative deployment work n*ln(n); this is a chosen growth shape, not measured cost.
Equation: n * _np.log(n).
model_counts
engine.scenario_types.OperationalLoadScenario.model_counts()Integer model-count grid from one through max_models.
Equation: _np.arange(1, max_models + 1).
total
engine.scenario_types.OperationalLoadScenario.total(n)Sum of alert, deployment, and conflict work in the scenario’s common capacity units.
Equation: alerts(n) + deployments(n) + conflicts(n).
alerts_per_model
Type: float.
Declared default: required.
conflict_coefficient
Type: float.
Declared default: required.
team_capacity
Type: float.
Declared default: required.
max_models
Type: int.
Declared default: required.
crossing_models
engine.scenario_types.OperationalLoadScenario.crossing_models
First grid count selected by the capacity-crossing mask; callers must ensure a crossing exists.
Equation: n = model_counts(); int(n[int(_np.argmax(total(n) >= team_capacity))]).