QC Samples

With the QC Samples form (study > Volumetrics > Volumetric Study > QC Samples) you can do two things:

QC the sampled values as generated with 'Tornado' or 'Latin Hypercube' and which are based on the probability density functions (PDFs) of each of the uncertainty parameters.
Manually add realizations and parameter values in case of sampling method 'Ad hoc'.

Each uncertainty parameter requires a value per realization, after which the computational model can be run with a single value for each parameter distribution. When your design contains correlated uncertainty parameters, you can only view the sampled probability percentages of the child uncertainty parameters on this form. The physical values for the children uncertainty parameters are calculated on the next form in the workflow, Implied QC Values.

Important For the uncertainty parameters Depth Multiplier or Thickness Multiplier, verify that the 'Multiplier distribution settings' entry field(s) on the model > 3D Structure > Depth or Thickness forms are set to 0 before sampling with Tornado. The reason is that the Realizations table adopts this value as the Multiplier method's reference value. Make sure to resample the Designs form in case you updated the value.

To add realizations with Ad hoc

Under Designs, select the design with the Ad hoc sampling technique.
Click on the 'Add realization' icon at the upper-right side of the table to add a realization to the table.

For a new 'Ad hoc' design, the Realizations table is empty at first. In case the Realizations table is already filled with data/samples, this means you are resampling a design (with Ad hoc) that was sampled previously. In this case, it is recommended to remove the existing samples (multiselect of rows in the table is enabled) and start entering ad hoc samples afresh.

In the 'Concept' column, select the volumetric concept that contains the uncertainty parameter(s) for which you want to enter sampling value(s). Upon selection, for each active uncertainty parameter (i.e. uncertainty parameters that are selected on the Designs form) a column is generated in the Realizations table.
Enter a value for each uncertainty parameter in the Realizations table by typing a value in the corresponding cell. Double click on a cell to make it editable.

(Not for seed parameters) For ad hoc realizations, you can optionally use the Autofill Realizations button () on the form to fill in the minimum, most likely, maximum or reference values as defined in the PDF on the Uncertainty Parameter dialog without having to look it up. To do this, select a realization (multiselect is supported) in the table. Click the Autofill Realizations button, select one of the values from the drop-down on the Autofill Samples dialog and click OK. The chosen value will be filled in for the selected realization(s) in the Realizations table.

When the selected display for the table is Probability and you use Autofill Realizations, it will autofill the probability associated to the min, most likely, max and reference values as defined in the PDF of the uncertainty parameter on the Uncertainty Parameter dialog. For example, when you select min as a reference value, the probability will be 100%; when you select most likely as reference value, the probability will be 50% (in case of non-skewed distribution), etc.

You can also copy-paste from a Microsoft Excel™ spreadsheet.
Grayed-out cells belong to another concept and should stay empty while not being used.

Important Make sure that when entering ad hoc sampling values, the values fall within the parameter's uncertainty range as specified on the parameter's Uncertainty Parameter dialog.

Enter multiple realizations by repeating the previous steps. To add multiple rows at once, click the 'Add multiple realizations' icon at the upper-right side of the table. The other options in the toolbar are: duplicate a row (), or delete a row ().
When you have finalized adding realizations and values, click Apply and keep the form open, or click OK to move to the next workflow step. Depending on the uncertainty parameters in your study, this can be Implied QC Values (when you have correlated uncertainty parameters) or Run Study (when you only have independent uncertainty parameters in your study).

The modified ad hoc designs are indicated in italics in the Designs list. If your study consists of multiple ad hoc designs with one or more invalid realizations, those designs are also marked invalid. You can identify invalid designs with error icon and a tooltip explaining the invalidity in the design and the realization(s). In this case, upon clicking Apply or OK an error message is shown and you cannot proceed to the next workflow step. However, if you then click Cancel or close the form, the valid designs are saved while the invalid designs will be removed when re-opening the form.

Sync Ad Hoc Values () Click this button if you made changes to the PDFs or dependency settings of uncertainty parameters. Upon clicking, the values of all ad hoc designs are updated based on the latest settings (except for ‘Custom’ values). When uncertainty parameters are correlated (parent/child relationship), it depends on the type of update (e.g. distribution, reference value, correlation factor) in combination with the type of correlation (i.e. partial or full dependency) whether the child values are impacted, or not.

In general it can be stated that when you made an update to the distribution and/or reference value of a parent and/or child uncertainty parameter, you have to click the button to make sure the values on ‘QC Samples’ and ‘Implied QC Values’ forms are in sync with the latest settings.

The realizations table

The Realizations table at the right-hand side of the form contains the sampling results:

Table toolbar (depending on selected design, toolbar options vary) -

	Displays the values which are sampled from the uncertainty parameter's probability density function.
	Displays the cumulative probability of the sampled values (cumulative probability according to the exceedance principle, i.e. the probability that the value will be equaled or exceeded). This column stays empty when the uncertainty parameter is a seed.
	Add a new row to the table.
	Add multiple rows to the table.
	Duplicate the selected row.
	Delete the selected row.

Table rows - Each table row represents a realization that you generated by setting an uncertainty parameter to active on the Designs form, or by clicking the plus icon at the top-right of the table when the selected sampling method is Ad hoc sampling. Each row represents one realization in your volumetric calculation.

The first realization of each concept is the reference case, printed in bold.

Table columns - The active independent and partially correlated uncertainty parameters included in your uncertainty calculation. Uncertainty parameters that you defined in the various workflows, but which are not set to 'active' on the Designs form, do not appear as a column in the table. Their reference values are used.

Partially correlated uncertainty parameters are sampled in their CDF space, and always listed as percentage. The physical values of partially correlated parameters is calculated in the Implied QC Values step of the Volumetrics Study workflow.