Tunny Icon
TunnyDocs

The next-gen Grasshopper optimization tool.

Method Map

When you know what you want to ask but not which method answers it, the following overview may help.

Parameter importance

flowchart LR
    imp["Parameter importance"] --> impA["quick check<br/>Spearman / Ridge"]
    imp --> impB["accurate check<br/>Sobol / tree-based / SHAP"]

Distributions and relationships

flowchart LR
    dist["Distributions and relationships"] --> distA["one variable's distribution<br/>Histogram"]
    dist --> distB["compare distributions<br/>Box Plot"]
    dist --> distC["all pairwise correlations<br/>Correlation Matrix"]
    distC --> distD["inspect interesting pairs<br/>Scatter Matrix"]

Multi-objective convergence

flowchart LR
    conv["Multi-objective convergence"] --> convA["Convergence Indicators chart<br/>HV / IGD+ / ε / R2"]

Pick good trials

flowchart LR
    pick["Pick good trials"] --> pickA["holistic score<br/>TOPSIS / VIKOR / PROMETHEE"]
    pick --> pickB["full trade-off<br/>Pareto Front"]
    pick --> pickC["compare the picked few<br/>Radar Comparison + Comparison Table"]
    pick --> pickD["stability under input scatter<br/>Robustness"]

Structure in the design space

flowchart LR
    struct["Structure in the design space"] --> structA["topology-preserving 2D map<br/>SOM"]
    struct --> structB["variance plane<br/>PCA Biplot"]
    struct --> structC["nested grouping<br/>Hierarchical Clustering"]

Parameter–objective relationships

flowchart LR
    rel["Parameter–objective relationships"] --> relA["1 parameter<br/>1D PDP / Slice Chart"]
    rel --> relB["2 parameters<br/>2D PDP on a surrogate"]
    rel --> relC["observed values only<br/>Observed Contour"]
    rel --> relD["local landscape<br/>Response Surface 3D"]