variable_influence module¶
Plot the partial sensitivity of the functions.
- class gemseo.post.variable_influence.VariableInfluence(opt_problem)[source]
Bases:
OptPostProcessor
First order variable influence analysis.
This post-processing computes \(\frac{\partial f(x)}{\partial x_i}\left(x_i^* - x_i^{(0)}\right)\) where \(x_i^{(0)}\) is the initial value of the variable and \(x_i^*\) is the optimal value of the variable.
Options of the plot method are:
proportion of the total sensitivity to use as a threshold to filter the variables,
the use of a logarithmic scale,
the possibility to save the indices of the influential variables indices in a NumPy file.
- Parameters:
opt_problem (OptimizationProblem) – The optimization problem to be post-processed.
- Raises:
ValueError – If the JSON grammar file for the options of the post-processor does not exist.
- DEFAULT_FIG_SIZE = (20.0, 5.0)
The default width and height of the figure, in inches.
- database: Database
The database generated by the optimization problem.
- materials_for_plotting: dict[Any, Any]
The materials to eventually rebuild the plot in another framework.
- opt_problem: OptimizationProblem
The optimization problem.