AUTOCR Process

To access this process:

  • Sample Analysis ribbon >> Geochemical Processes >> Auto Correlation Analysis.
  • View the Find Command screen, select AUTOCR and click Run.
  • Enter "AUTOCR" into the Command Line and press <ENTER>.

See this process in the Command Table.

Process Overview

Auto-correlation analysis is used to quantify and define anomalous thresholds and halo size on regularly grided soil sample lines. AUTOCR can also be used to measure dispersion limits in stream sediments, as long as fixed distances are used for sampling.

The process calculates the auto correlation function R-L of a single field against sample DISTANCE or LAG (see Figure 1). By default, lag distance is used to calculate the auto correlation function. If sample distance, or a multiple of sample distance is required, use the parameter SAMPDIST.

Anomalous samples related to lag or sample distance (LAG or DISTANCE) are identified by strong and well defined peaks of R-L, the covariance between neighboring samples. Dispersion limits from mineralized stream sediments are defined by breaks in slope along the stream sediment train as defined by LAG-L.

Note: There is a restriction of 4000 samples for a given line or dispersion train. For the analysis to be valid the data points must be equi-distant from each other. All fields containing the same value should be removed prior to the analysis. A plot macro should be written to display the results.

File Handling

The input file (&IN) must contain a sample identifier (*SAMPID) which is declared on input. There is an obligatory output file (&OUT) of the results containing the fields LAG-L, DISTANCE, R-L and the significance levels for each observation which can be viewed by the plotting processes, e.g. PLOTDA and PLOTAN.

Input Files

Name

Description

I/O Status

Required

Type

IN

Input file. Must contain sample identity field.

Input

Yes

Undefined

Output Files

Name

I/O Status

Required

Type

Description

OUT

Output

Yes

Undefined

Output file includes LAG-L, DISTANCE. R-L the auto correlation function and SIGNIC the significance of the auto correlation function for use in graphical processes.

Fields

Name

Description

Source

Required

Type

Default

SAMPID

Compulsory sample identifier field.

IN

Yes

Undefined

Undefined

F1

First variable for evaluation. If no variables are selected all variables will be processed.

IN

No

Numeric

Undefined

F2

Second variable for evaluation.

IN

No

Numeric

Undefined

F3

Third variable for evaluation.

IN

No

Numeric

Undefined

F4

Fourth variable for evaluation.

IN

No

Numeric

Undefined

F5

Fifth variable for evaluation.

IN

No

Numeric

Undefined

F6

Six variable for evaluation.

IN

No

Numeric

Undefined

F7

Seventh variable for evaluation.

IN

No

Numeric

Undefined

F8

Eighth variable for evaluation.

IN

No

Numeric

Undefined

F9

Ninth variable for evaluation.

IN

No

Numeric

Undefined

F10

Tenth variable for evaluation.

IN

No

Numeric

Undefined

Parameters

Name

Description

Required

Default

Range

Values

SAMPDIST

 

Option

Description

(0)

Distance between sample points to calculate the auto-correlation function. If no distance is specified the sample distance is lag distance.

No

0

Undefined

Undefined

PRINT

>0 Display results on the screen (0).

No

0

0,1

0,1

Example

!AUTOCR &IN(L10),&OUT(L10ACR),@SAMPID='ID',*F1(BA),*F2(SR),@SAMPDIST=20

Error and Warning Messages

Message

Description

*** Warning *** nnnnn Samples exceed the maximum of 4000.

The first 4000 will be processed.

*** Error *** No numeric fields on file (filename).

>>> ERR 122 <<< ( fileno.) IN AUTOCR

 

*** Error *** SAMPID field (fieldname) is not in input file.

*** Error *** in reading data from file (filename).

>>> ERR 200 <<< ( fileno.) IN AUTOCR

 

*** Error *** Compulsory file, field or parameter missing.

>>> ERR 120 <<< ( fileno.) IN AUTOCR