Discussion of Control Parameters

This page is to use in brainstorming how to handle control parameters in PINT. These are things in .tim files that are not just TOAs, and things in .par files that are not actual parameters of the fit. These tend to break some of the nice modularity that PINT espouses, so we need to come up with a good way to handle them that is both clean and as backwards compatible as possible. To begin, let's document the control parameters that might need to be supported and split them up by how they are handled:

• Timing Parameters in .tim files

• JUMP

• Parameters in the .tim file that affect the fitter

• MODE (only in deprecated TEMPO usage)

• PHASE

• Parameters that affect TOAs in .tim files. These affect only TOAs and do not need to be considered by the model or fitter.

• EQUAD

• EFAC

• GLOBAL_EFAC

• EMIN

• EMAX

• EFLOOR

• T2EFAC

• T2EQUAD

• INFO

• END

• INCLUDE

• SKIP/NOSKIP

• FMIN

• FMAX

• TIME

• SIGMA

• TRACK (deprecated in .tim files, I believe)

• Parameters in the .par file that affect how TOAs are interpreted or converted

• EPHEM

• EPHVER

• CLK

• T2CMETHOD

• TIMEEPH

• Parameters in the .par file that are really just informational output

• START/FINISH (if they don't have a '1' appended)

• TRES

• NTOA

• Parameters in the .par file that affect how the model is defined, but aren't model parameters

• BINARY

• UNITS (or should this be in a different category?)

• Parameters in the .par file that affect the operation of the fitter (or how the model is computed)

• NITS

• MODE

• ECORR

• SOLARN0

• ECL

• RNAMP/RNIDX (you might consider these 'model' parameters, but since they aren't part of a deterministic model but something that the fitter must use, I'm putting them here)

• START/FINISH (if they are activated with a '1' appended)

• TRACK

• DIALATEFREQ

• PLANET_SHAPIRO

• CORRECT_TROPOSPHERE

• JUMP