Process Help

KNA - optimise Kriging parameters for use in grade estimation

 

Process Name

Menu Path

Link to Command Table

KNA

Access using command line

Used by Advanced Estimation wizard

Click here

 

Introduction

KNA (Kriging Neighbourhood Analysis) is used to optimize the parameters for grade estimation. Typically multiple estimations are performed on a small number of blocks and the results are averaged for each estimation and compared to find an optimal set of parameters. The parameters are then used by the COKRIG process to estimate grades into a block model by a range of methods including Kriging.

The input to KNA is similar to COKRIG:

  • SAMPLES: Samples file

  • TESTBLKS: Up to three sets of model blocks to be tested are identified by their XC, YC and ZC coordinates.  Each set can represent a different area and is identified by the BLKGROUP field

  • EPAR: Specifies estimation parameters such as search volume identifier, kriging method, block size and discretization points

  • FIELDS: Input (grade) and output (estimation/KNA stats) field names for each estimation

  • VMODEL: Contains the variogram model parameters

  • SPAR: Contains the search volume parameters referenced in the EPAR file

The output file OUT lists block size, block group, discretization, search parameters and KNA statistics.

 

Files, Fields and Parameters

Input Files

Name

Description

I/O Status

Required

Type

SAMPLES

A file containing sample positional information and supporting attributes.

Input

Yes

Undefined

TESTBLKS

A file containing centre positions of test blocks to perform grade estimation on. It can optionally contain a numeric field to specify groups of blocks to calculate statistics per group which must be specified as the BLKGROUP field.

Input

Yes

Undefined

EPAR

The input estimation parameter file used to specify parameters and for each estimation.

This must contain the following fields:

  • SREFNUM
  • IMETHOD
  • DISCX
  • DISCY
  • DISCZ
  • BLKSZX
  • BLKSZY
  • BLKSZZ

SREFNUM defines the reference of the search parameters found in the SPAR file.

 IMETHOD is used to specify the Interpolation method

  • 3=Ordinary Kriging (default)
  • 4=Simple Kriging using sample means
  • 10=Simple Kriging using means defined in FIELDS
  • 11=Simple Kriging using local means defined in the prototype).

DISCX, DISCY, DISCZ specify the number of discretization points.  

BLKSZX, BLKSZY, BLKSZZ specify the dimensions of the cells to be interpolated into.

Input

Yes

Undefined

FIELDS

A file that contains field names of variables to be used in estimation. Input variables must be included under the mandatory column IN_VAR and each of these fields must be present in the SAMPLES and VGRAM file.

If more than 1 variable is supplied Multivariate (Co)Kriging will be performed. If IMETHOD=10 is used, the column SKMEAN must be used to specify the mean per variable.

If IMETHOD=11 is used the column LOC_MEAN must be used to specify the local mean fields in the prototype model.

Input

Yes

Undefined

VMODEL

The input (cross-)variogram model parameter file. If more than 1 variable is suppled in the FIELDS file (i.e. multivariate estimation), this file must contain the columns GRADE and GRADE2.

Input

Yes

Undefined

SPAR

The input search parameter file. This must contain the following 12 mandatory fields:

  • SREFNUM which is used to store a reference number for each record which is then specified by the SREFNUM parameter in COKRIG.
  • SDIST1, SDIST2, SDIST3 specifying the search distances in the X, Y and Z directions respectively.
  • SAXIS1, SAXIS2, SAXIS3 specifying the first, second and third axes which the search volume is to be rotated around (1 = x, 2 = Y, 3 = Z).
  • SANGLE1, SANGLE2, SANGLE3 which specify the clockwise angles which search volume is rotated around the axes specified by SAXIS1, SAXIS2, SAXIS3.
  • MINNUM1 which is the minimum number of samples required per estimate
  • MAXNUM1 which is the optimum number of samples to be used per estimate.

Input

Yes

Undefined

Output Files

Name

I/O Status

Required

Type

Description

OUT

Output

No

Undefined

The output file which will contain statistics for each estimation which has been performed.

Fields

Name

Description

Source

Required

Type

Default

XPT/YPT/ZPT

X/Y/Z coordinate of sample data in the SAMPLE file.

IN

Yes

Undefined

XPT/YPT/ZPT

KEY

Key field used to specify the field limiting the number of samples for estimation using the optional OPTKEY and MAXKEY parameters in the SPAR file. The field must exist in the SAMPLES file.

IN

No

Undefined

Undefined

BLKGROUP

Numeric field in the PROTO file used to split test blocks into groups to calculate statistics from.

IN

No

Undefined

Undefined

Parameters

Name

Description

Required

Default

Range

Values

VSETNUM

The reference number of the variogram model set to be used from the VMODEL file

Yes

1

Undefined

Undefined

NOKSTATS

Set this parameter to 1 to prevent output of Kriging statistics. Typically used with BLKCOV=1 when testing the number of discretization points to be used.

No

0

0,1

0,1

BLKCOV

Set this parameter to 1 to calculate and output block covariance. Typically used for testing discretization.

No

0

0,1

0,1

NBLKCOV

The number of random samples to use for calculating block covariance per KNA run.

A value of at least 20 is recommended to ensure reasonable precision.

No

20

Undefined

Undefined

PRNT

The level of detail for text printed to the command window. A value of 0 only prints errors, a value of 1 additionally prints warnings and progress, a value of 2 additionally prints further information.

Yes

0

0,1

0,1

NTHREADS

Number of threads to be used for the main calculation. Any value less than 1 will automatically select the values based on the number of virtual cores on the computer.

No

-1

Undefined

Undefined