Scan proposers for strategy generation.

Introduction to the strategy modules

There are three functions of a data collection strategy that can be clearly separated.

• The "target": What do we want to measure? Are we satisfied with the current scans, or do we need more?

• The "limitations". Where does the hardware allow us to measure? Can we avoid a long unit cell axis? Can we avoid less favorable positions for the cryostat?

• The "propositions". Think of some scans that we could do to measure more data.

For all three a class-hierarchy is present in the library. To make a complete data collection strategy, one target object, one limit object and one proposition object must be combined.

For non-data collection strategies currently a different concept is being used, consisting of only one class hierarchy of "strategies". This might change any moment. However, these classes are not so important for further development, because they are only present for GUI use and normal scripting applications will do fine without them.

Usage Example

This is a piece of code from the strategy code in the "Collect" GUI:

# Build limiter
from strategy import limit
lim=limit.CompositeLimit()
lim.Add(limit.HardwareLimit(self.experiment.hardware))
if self.experiment.efficlimit:
lim.Add(limit.EfficiencyLimit(thetamax=self.experiment.dataset.ThetaMax()))
if self.experiment.lowtemplimit:
lim.Add(limit.CoolerLimit())
if self.experiment.longaxislimit:
vec=self.experiment.cell.ShortestReciprocalLatticeVector()
lim.Add(limit.VectorAvoid(vec))
# Build target function
from strategy import target
if self.experiment.redundancy<1.0:
tar=target.RedundantTarget(dataset=self.experiment.dataset,
completeness=self.experiment.completeness,
limiter=lim,redundancy=self.experiment.redundancy)
else:
tar=target.OnceTarget(dataset=self.experiment.dataset,
completeness=self.experiment.completeness,
limiter=lim)
#
pro=self.experiment.strategytype(
target=tar,
hardware=self.experiment.hardware,
cell=self.experiment.cell,
dx=self.experiment.dx)
pro.Calculate()
return pro.ScanSet()

Variables defined by the proposer module.

a single variable proposerlist is defined by this module. It refers to a list of all the proposer classes defined. This can be used in a GUI to make user selection of the data collection strategy possible,

For this purpose, each proposer class has a 1-line doc-string that uniquely describes its functionality.

normang ( p )