BM Stepwise Gaussian

The BM Stepwise Gaussian option allows you to transform two or three block model variables such that the bivariate (or trivariate) and marginal distributions of the transformed variables are all standard Gaussian. This means that each one of the variables are transformed to standard Gaussian in such a way that the correlation between them is zero. This transformation is useful when simulating or estimating variables that are correlated to each other. By applying this transformation, the Gaussian variables have a zero correlation coefficient between them, allowing you to treat them separately.

Note:  Simulating the correlated variables can lead to a loss of the original correlation between the variables in the resulting model.

This option can also be accessed by selecting the BM Stepwise Gaussian button from the Gaussian Transformations toolbar.


Instructions

On the Block menu, point to Gaussian Transformations, then click BM Stepwise Gaussian.

Settings

Follow these steps:

  1. Enter a name for the Specification file, or select it from the drop-down list. The drop-down list displays all files found in the current working directory that have the (.bmn) extension. Click the Browse icon to select a file from another location.

  2. Select a Scenario ID. To create a new ID, click the New icon as shown below, and provide a unique name for the current panel settings.

  3. Select the Block model from the drop-down list. Click the Browse icon to select a file from another location.

  4. Select the original primary grade variable from the Mean (X Axis) drop-down list.

  5. Select the grade variable that will be used to store the Gaussian transformation of the primary variable from the drop-down list labelled Stepwise conditional mean.

  6. Select the grade variables that will be used to store the original values of the Residual 1 (Y Axis) and Residual 2 (Z Axis) conditioned variables.

  7. Select the grade variables that will be used to store the Gaussian transformation of the residual variables from the Stepwise conditional residual drop-down fields.

    Note:  Using a second residual variable in steps 7 and 8 is optional and is used if you wish to conduct a trivariate study.

  8. Select Apply logarithm constant if you want to apply the base logarithm function to all values.

    Important:  In order for the logarithm to be defined all original values must be positive.

    The specified logarithm constant will be added to the calculated logarithm.

  9. Select Cut samples if you want to apply cut-offs to the values used in the transformation. You will need to specify a lower grade cut value (grades lower than this value are set to this value) as well as an upper grade cut value (grades above this value are set to this value).

Reference Distribution

Use this pane to setup a different dataset to create the transformation lookup table to use with the block model data. In this case all the transformation conversion is defined using the reference database, but the data that gets transformed is defined through the Samples Database section of the interface.

Follow these steps:

  1. Select Use Reference Distribution to enable the options on this pane.

  1. Select either a database or map file as your sample source.

    • To select a database, enable the option ISIS file, then select the file from the drop-down list. Click folder icon to select a file from another location.

      You can also select an ODBC link for database files found on site servers.

    • To select a map file, enable the option labelled Use Map File, then select the file from the drop-down list. Click the Browse button to select a file from another location.

  1. Map the correct fields by filling out the Samples Fields information.

    1. Sample group - Enter the name of the groups (database keys) to be loaded in the field. Wildcards (* multi-character wildcard and % single character wildcard) may be used to select multiple groups. Multiple groups only apply to Isis databases (ASCII map files consist of one group).

    2. Select the field containing the Reference distribution.

    3. Select the field that will hold the Weights.Sample values are not changed by the weighting, only the relative importance in the distribution is adjusted. Leave this field blank if you do not want to apply any weighting.

  2. Select the grade variables that will be used to store the original values of the Residual 1 (Y Axis) and Residual 2 (Z Axis) conditioned variables.

    Note:  Using a second residual variable in steps 7 and 8 is optional and is used if you wish to conduct a trivariate study.

Samples Filter

Use this pane to include any restrictions to your data by using the four specialised filters.

Follow these steps:

  1. Include any restrictions to your data by using the four specialised filters.

Block Selection

Use this pane to limit the blocks that will be used. All of the selections on this pane are optional.

Options available:


Click Apply and Run to begin the simulation. This will save your settings in to the specification file and start the simulation run.

Click OK to save your settings to the specification file without running the simulation.

Click Cancel to close the panel without saving any settings.