pargrid¶

The module provides tools for creating grids to scan over parameters.

It supports range, linspace, logspace and geomspace which are similar to their analogues from numpy. It also supports a list of values passed from the command line.

Positional arguments

name – define name if the parameter grid

Options

--range – define grid via NumPy arange (endpoint not included)

Positional arguments: PARAMETER START STOP STEP

--logspace – define grid via NumPy logspace (endpoint included)

Positional arguments: PARAMETER START_POWER STOP_POWER NUMBER

--geomspace – define grid via NumPy geomspace (endpoint included)

Positional arguments: PARAMETER START STOP NUMBER

--linspace – define grid via NumPy linspace (endpoint included)

Positional arguments: PARAMETER START STOP NUMBER

--list – define grid via NumPy array of values

Positional arguments: PARAMETER SPACE SEPARATED VALUES (VALUE_1 VALUE_2 VALUE_3…)

--segments – define a segmentation of grid to allow parallel scanning of large grids

Positional arguments: NUMBER OF SEGMENTS CURRENT SEGMENT

-v, --verbose – define verbosity level

Examples

Generate a linear grid for the parameter ‘E0’:

./gna \
    -- gaussianpeak --name peak \
    -- pargrid scangrid --linspace peak.E0 0.5 4.5 10 -vv

Provide a list of grid values from a command line:

./gna \
    -- gaussianpeak --name peak \
    -- pargrid scangrid --linspace peak.E0 1 2 8 -vv