UNIFIT issueshttps://iffgit.fz-juelich.de/zorn/unifit/-/issues2020-07-08T13:07:06Zhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/23Simplify scripting2020-07-08T13:07:06ZReiner ZornSimplify scriptingThe function of the shell script can be integrated into the python script `unifit_p1`. This would make the shell script unnecessary simplifying the whole structre and removing a source of errors for future extensions. This structure would also make the implementation of `def` easier (issue #21 ).The function of the shell script can be integrated into the python script `unifit_p1`. This would make the shell script unnecessary simplifying the whole structre and removing a source of errors for future extensions. This structure would also make the implementation of `def` easier (issue #21 ).Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/21Low-level macro definitions2020-07-08T13:07:06ZReiner ZornLow-level macro definitionsEspecially in the data section it is often necessary to repeat sequences of commands near-identically, e.g.
file PEMA_400_60.sqt
set Temp 400
set Q 0.6
group Qs t400 0.8 1
file PEMA_420_60.sqt
set Temp 420
set Q 0.6
group Qs t420 0.8 1
file PEMA_440_60.sqt
set Temp 440
set Q 0.6
group Qs t440 0.8 1
file PEMA_440_20.sqt
set Temp 440
set Q 0.2
group Qs t440 0.8 1
...
It would be nice if one could first define a macro, like
def pfile $1 $2
file PEMA_$1_$2.sqt
set Temp $1
set Q 0.$2
group Qs t$1 0.8 1
end
and then just use the user-defined command:
pfile 400 60
pfile 420 60
pfile 440 60
pfile 440 20
...Especially in the data section it is often necessary to repeat sequences of commands near-identically, e.g.
file PEMA_400_60.sqt
set Temp 400
set Q 0.6
group Qs t400 0.8 1
file PEMA_420_60.sqt
set Temp 420
set Q 0.6
group Qs t420 0.8 1
file PEMA_440_60.sqt
set Temp 440
set Q 0.6
group Qs t440 0.8 1
file PEMA_440_20.sqt
set Temp 440
set Q 0.2
group Qs t440 0.8 1
...
It would be nice if one could first define a macro, like
def pfile $1 $2
file PEMA_$1_$2.sqt
set Temp $1
set Q 0.$2
group Qs t$1 0.8 1
end
and then just use the user-defined command:
pfile 400 60
pfile 420 60
pfile 440 60
pfile 440 20
...Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/19Increase number of digits in unifit_res.txt2020-06-10T06:38:28ZReiner ZornIncrease number of digits in unifit_res.txthttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/17Plot unfitted data points2020-05-18T06:28:13ZReiner ZornPlot unfitted data points...maybe in a different color, size......maybe in a different color, size...Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/14Create 'for' command for calculate, theory, and plot2020-05-14T16:07:23ZReiner ZornCreate 'for' command for calculate, theory, and plotPutting `for 5` in front of `calculate`, `theory`, or `plot` shall restrict the command to file 5, putting `for 1 4` to files 1..4.Putting `for 5` in front of `calculate`, `theory`, or `plot` shall restrict the command to file 5, putting `for 1 4` to files 1..4.Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/13Improve rplot2020-05-14T15:21:28ZReiner ZornImprove rplot* [ ] Enable resolution plot also if fit function is not calculated (not: Function calculation not requested -> no plot.)
* [ ] Option to plot vs point number instead of x.* [ ] Enable resolution plot also if fit function is not calculated (not: Function calculation not requested -> no plot.)
* [ ] Option to plot vs point number instead of x.Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/10For 'numerical' convolution include delta functions in theory plot2020-05-14T04:05:47ZReiner ZornFor 'numerical' convolution include delta functions in theory plotReiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/8Normal MC error analysis2020-05-11T16:41:52ZReiner ZornNormal MC error analysisIn a similar way as bootstrap error calculation is implemented, the easier Monte Carlo error calculation should be implemented. There should be a choice between the data itself and its model representation as a starting point for generating the counterfactual data.In a similar way as bootstrap error calculation is implemented, the easier Monte Carlo error calculation should be implemented. There should be a choice between the data itself and its model representation as a starting point for generating the counterfactual data.Reiner ZornReiner Zornhttps://iffgit.fz-juelich.de/zorn/unifit/-/issues/7= and + do not work for group parameters2020-05-11T16:32:57ZReiner Zorn= and + do not work for group parametersMost users will not work with enumerated groups anymore. But those who do will run into problems when using `+` to start the next group or `=` to stay with the same. Also, `=` should also work for the new standard, labelled groups.Most users will not work with enumerated groups anymore. But those who do will run into problems when using `+` to start the next group or `=` to stay with the same. Also, `=` should also work for the new standard, labelled groups.https://iffgit.fz-juelich.de/zorn/unifit/-/issues/6replace HSL routines by true open source if possible2020-05-11T16:07:45ZReiner Zornreplace HSL routines by true open source if possibleMost of the Harwell Scientific Library routines can be replaced by true open source code. This is not possible for VA05 because of its connection with FZJ-written routine KOVA. But at least four downloads from HSL could be avoided.Most of the Harwell Scientific Library routines can be replaced by true open source code. This is not possible for VA05 because of its connection with FZJ-written routine KOVA. But at least four downloads from HSL could be avoided.Reiner ZornReiner Zorn