Sample parameter points – prof-sampleparams¶
prof-sampleparams samples parameter points either randomly to use them as anchor points for MC data production or along straight lines for line-scan validation with prof-scanchi2.
In the random mode (default) the points are sampled randomly from a hyper-cube. The path to the range file must be given as first argument. Alternatively, two arguments are accepted with files containing the lower and upper parameter bounds. The number of points is controlled with -N NUMRUNS. The parameter points are written to files in numbered directories under OUTDIR (see -o):
OUTDIR/000/used_params 001/used_params ...
Line scan mode¶
Line scan mode can be activated with -s. In line scan mode equally distant points are generated along a straight line. prof-scanchi2 compares the MC data at this points with the prediction of polynomial parameterisations.
The points are only sampled from the ranges given by the first argument. These ranges should be the same that were used for the polynomials with prof-interpolate, or smaller, because otherwise the polynomials are evaluated in regions of extrapolation. Alternatively, the ranges can be specified by two parameter point files.
Sample 100 points randomly from the ranges in generator.ranges and create the files under mcruns/:
prof-sampleparams -N 100 -o mcruns generator.ranges
Sample 10 points along a straight line in the ranges of generator.ranges. The line is rooted at the parameter values of the first tune in the results file results.pkl and is oriented in the direction of the largest uncertainty:
prof-sampleparams -s -N 10 -o linescan --scancenter results.pkl,0 --direction shallow generator.ranges
The --datadir DATADIR and related options are used as normal to specify the reference data, MC runs, and interpolation objects: see the path options page.
Line scan mode¶