This is an old revision of the document!
Overview
Tables allow simulation data to be presented in a tabular form. The correlation table is used after a sensitivity analysis to present correlation coefficients between model inputs and model outputs. The table is only available after a probabilistic simulation or a sensitivity analysis has been performed.
For a correlation table, the following dimensions are available:
Outputs - The outputs that have been selected for the table.
Time - If any of the outputs is time dependent, then the time dimension allow you to select which time points to show data for. It is also possible to show the minimum/maximum value across the time vector and the time points for which the minimum or maximum occurs.
Indices - For each dimension of the selected outputs, a dimension is available for the table data. These dimensions allow you to select all, or only a sub-set of the indices of an
index list.
Correlation type - The coefficients to calculate.
Inputs - Whether to include all inputs or only inputs that can influence the selected objects. This dimension can also include an R2 value.
Transformation - How to transform the data before the coefficients are calculated (none, rank or log)
Creating a correlation table
Table controls
The controls beneath the table let you change the appearance of the table.
Highlight
Highest - Select to highlight the highest value in each column.
Lowest - Select to highlight the lowest value in each column.
Condition - Select to enter a condition. The condition is on the form xstart < x < end and will be evaluated for each cell in the table. Cells who’s value fulfills the condition will be highlighted.
Editing a correlation table
To edit a correlation table either right-click the table in the tables window and select Edit, or double-click the table, or click the button in the title bar.
The table editor has the following tabs:
Outputs - Select which outputs to include in the table. The outputs should all have the same dimension.
Indices - Depending on the dimension of the outputs that were selected, you can select which indices to include or to include all indices.
Time - If the outputs are time dependent, this tab allows you to select which time points to display data for.
Specific - Select either all time points or specific time points to include
First - Display the first time point
Last - Display the last time point
Max over time - Display the maximum value for each output over time
Time for max - Display the time point for the maximum value for each output
Min over time - Display the minimum value for each output over time
Time for min - Display the time point for the minimum value for each output
Correlation - This tab allows you to select which coefficients to calculate and how to transform the data.
Correlation - Select which coefficients to calculate
Transformation - Select how to transform the data before coefficients are calculated. Note that transformation is sometimes already performed by the selected correlation. For instance,
Spearman is the same as
Pearson with ranked data.
Inputs - Select whether to include all selected parameters, or to filter out parameters that cannot influence the selected outputs.
-
See also