Calibration Sheet¶

This sheet contains calibration data including enumerations. It has the following columns:

calibrator name (required)

Name of the calibration. Used as a reference in the Parameters or Commands sheet.

type (required)

One of polynomial, spline, enumeration, java-expression or time.

Detailed in sections below.

time for converting a raw integer or float value into a timestamp value.

calib1

Contents depends on the chosen type. See sections below.

calib2

Contents depends on the chosen type. See sections below.

Polynomials¶

If the type is set to polynomial, polynomial calibration is performed.

calib1 (required): List the coefficients, one per row starting with the constant and up to the highest grade. There is no limit in the number of coefficients (i.e. order of polynomial).
calib2: (not used)

Note that the polynomial calibration is performed with double precision floating point numbers even though the input and/or output may be 32 bit.

Splines¶

If the type is set to spline, linear spline (pointpair) interpolation is performed. As with polynomial calibration, the computation is performed with double precision numbers.

calib1 (required): Start point: x from (x, y) pair.
calib2 (required): Stop point: y from (x, y) pair.

Enumerations¶

If the type is set to enumeration, the calibrator can be used to map enumeration states.

calib1 (required): Numeric value
calib2 (required): Text state corresponding to calib1.

Java Expressions¶

The type java-expression serves as a catch-all. They can be used for float or integer calibrations.

calib1 (required)

The textual formula to be executed. This expression will be enclosed and compiled into a class like this:

package org.yamcs.mdb.jecf;
public class Expression665372494 implements org.yamcs.mdb.CalibratorProc {
    public double calibrate(double rv) {
            return <expression>;
    }
}

The expression should return a double, but Java will convert implicitly any other primitive type to a double.

Java statements cannot be used, however the ternary operator ? : can be used; for example this expression would compile fine:

rv > 0 ? rv + 5 : rv - 5

Static functions can be also referenced. In addition to the usual Java ones (e.g. Math.sin, Math.log, etc) user-own functions (available in the Java classpath) can be referenced by specifying the full class name:

my.very.complicated.calibrator.Execute(rv)

calib2

(not used)

Time¶

If the type is time, this calibrator allows to convert a raw integer or float value into a timestamp value by using the raw value as an offset from a well known epoch or from another parameter. Optionally allow to use an (offset:scale) which can be used to scale the raw value from other units (e.g. millseconds) to seconds.

Known epochs are GPS, TAI, UNIX and J2000.

The conversion is performed as follows:

When using a known epoch: engValue = <epoch>_yamcs_difference + offset+rawValue*scale.
When using another parameter p: engValue = p.engValue + offset+rawValue*scale.

calib1 (required)

Something of the shape epoch:<epoch> or parameter:<parameter reference>. The reference has to be to a parameter of type time.

calib2

Optionally something of the shape offset:scale where both offset and scale are numbers.

If unset, this defaults to 0:1