Filter Parameter Information in the Design Gateway
When viewing parameter information on the Parameters
tab in the Design Gateway,
you can filter the data to make it more useful to your design process.
From the Design Gateway,
select a model.
Select the component, and click the Parameters
tab.
Click the button above the parameter list.
The Table Filters dialog box appears.
Determine how you want to filter the parameters:
Option |
Description |
Name |
Select this option to filter the parameters
based on the parameter name.
You can use the “*” wildcard with the Name
filter.
For more information, see About Parameter Names. |
Display Name |
Select this option to filter parameters
based on the name that appears in visuals (e.g., graphs, correlation
maps, etc.). You can use the “*” wildcard with the Display Name
filter.
For more information, see About Display Names.
|
Mode |
Select this option to filter the parameters
based on the mode (Input, Output,
Input/Output, or Local).
For more information, see About the Parameter Mode.
|
Value |
Select this option to filter the parameters
based on the parameter values. |
Unit |
Select this option to filter the parameters
based on the unit. For more information, see About Units.
|
Type |
Select this option to filter the parameters
based on the type (Boolean, Integer,
Real, or String). |
Description |
Select this option to
filter the parameters based on the parameter description. You can use the “*” wildcard with the Description filter. |
Allowed Values |
Select this option to filter
the parameters based on the allowed values (the values that this parameter
can be set to when the corresponding component executes). For more information,
see About Allowed Values. |
Mapped |
Select this option to filter the parameters
based on whether or not they are mapped. |
Save to DB |
Select this option to filter the
parameters based on whether or not the values for the parameter are saved
to the results database. If this option is not selected, parameters that
are not saved will not be displayed on the Parameters or History
tabs in the Runtime Gateway
during and after execution. For more information, see Setting Gateway Preferences. |
Click OK to save your changes and to close the
Table Filter dialog box.
Isight
updates the Parameters tab based on your filter
selections. Any column selected as part of your filter options displays
the icon to the left of the column header. The text to the
right of the button shows you how many parameters (rows)
match your filter based on the total number of parameters (rows).
Perform any of the following actions, as desired:
-
Add to the existing filter by clicking the button.
-
Remove the defined filter, and return the parameter list to its original
state, by clicking the button.
Save the model, if desired.
Filter Parameter Information in the Runtime Gateway
You can use the Runtime Gateway
Data Filter to filter the history of runs for
the selected component, which reduces the amount of information displayed
and can allow you to better analyze the execution data. Filtering can
be based on the desired minimum/maximum parameter values included and/or
based on the “grading” of the runs using the problem formulation
(lower/upper bounds, objectives).
The filtered data set is displayed on the Parameters,
History, Data Analysis,
and Graphs tabs. However, postprocessing graphs
from the process
components (e.g., the Main Effects graphs for DOE) use the full, unfiltered
data set even after filtering.
From the Runtime Gateway,
access one of the following tabs:
- Parameters
- History
- Data Analysis
- Graphs
Verify that you are viewing the parameters for the desired component.
If not, select the correct component using the model explorer.
At the bottom of the tab, click Data Filter.
The filtering options appear.
The Data Filter options are divided by input
and output parameters. Parameter information is displayed in the Data
Filters tables near the bottom of the tab.
Change the Min and Max
values for the desired input and output parameters. The initial minimum
and maximum values are set from the range of the full data set. The data
set presented in the Runtime Gateway
for the selected component is dynamically altered after each change to
the minimum and maximum values. The number of data points currently presented
is displayed at the top of the Data Filters table.
Click Options, and perform any of the following
actions to further filter your data:
-
To control the filtering based on the minimum/maximum values of parameters
in the data set, use the following options:
Option |
Description |
Activate min/max filters |
Filters the data set
using specific minimum/maximum values of parameters
in the data set and selectively turns off/on the filtering
without losing the defined minimum/maximum values. |
Show active min/max only |
Displays only parameters
whose defined minimum/maximum values have resulted in data points being filtered
in the Data Filters tables. |
Clear all min/max filters |
Returns all min/max
values to the minimum/maximum values from the full data set (removing
all min/max filtering). |
-
To control the filtering based on the grading of the runs using the
problem formulation, use one of the following mutually exclusive options:
Option |
Description |
Show feasible and infeasible |
Shows runs whether
they are feasible (meet all defined lower/upper bounds) or infeasible
(violate at least one lower/upper bound). |
Show feasible only |
Shows only those runs
that do not violate any lower/upper bounds on parameters as defined in
the problem formulation. |
Show infeasible only |
Shows only those runs
that violate at least one lower/upper bound on a parameter as defined
in the problem formulation. |
-
Click Show improvements only to show only those
runs that are considered “improvements” over all previous runs with
respect to any defined objectives (and satisfaction of constraints) in
the problem formulation.
-
Click Show exact runs only to exclude all approximation
runs (if you used an approximation in your model).
-
Click Show Pareto Points to display
only the Pareto points in your design. By default, all points of the
selected component are displayed. Selecting this option filters out all
non-Pareto data points so that only the Pareto solutions are shown. Although
this option is available for all components, the computed Pareto set
is closest to the true Pareto set only when multi-objective optimization
techniques (i.e., AMGA, Multi-Objective Particle Swarm, NCGA, and NSGA-II)
are used.
-
Click Apply Filters to all Postprocessing to
control whether or not filters are used in all of the postprocessing
tools (including graphs, tables, and data analysis tools). By default,
this option is not enabled, and any filtering or feasibility options
that are in use are ignored when a postprocessing tool is used. Selecting
this option causes postprocessing tools (including computing a Pareto
set) to use only the data provided after using the filtering and feasibility
options.
Note:
This option is not available when using the Six Sigma
component’s Reliability Technique analysis type.
-
Click Filter by distance to control the filtering based on the Euclidean distance from a target.
The Euclidean distance is calculated in a space where the ranges for
all parameters are normalized to [0, 1] using the maximum and minimum
values from the history. - In the Filter by Distance dialog box, enter the
Filter Target values for the parameters. By default,
Isight
uses the target values set in the problem formulation.
- Enter the Number of rows to display. By default,
the value is 0. Leaving this value at 0 shows all the runs in the history.
When targets are set for filtering, a new column labeled Filter
Distance appears in the History table.
To sort the rows based on the filter distance, click the column heading.
The Filter Distance column is omitted when you
export the data to Excel or to a text file.
From the Runtime Gateway,
select View Run to view values from different
runs for the current job.
From the Runtime Gateway,
click Run Directory to select and view any file
in the associated working directory if any files were saved with the
job and you are running the job locally.
|