Ticket #5844 (closed: fixed)

Opened 8 years ago

Last modified 5 years ago

SANS: Fitting rescale and shift

Reported by: Anders Markvardsen Owned by: Gesner Passos
Priority: critical Milestone: Release 2.3
Component: MantidPlot Keywords:
Cc: richard.heenan@…, stephen.king@…, gesner.passos@… Blocked By:
Blocking: Tester: Anders Markvardsen

Description

Following on from #2159. Here allow rescale and shift to be fitted

Change History

comment:1 Changed 8 years ago by Anders Markvardsen

sans: allow fitting rescale and shift. re #5844

Changeset: 0ad65c1e3ffa6dca0e06f299e3d3aa5cea0e0258

comment:2 Changed 8 years ago by Anders Markvardsen

sans: allow fitting rescale and shift. re #5844

Changeset: 0ad65c1e3ffa6dca0e06f299e3d3aa5cea0e0258

comment:3 Changed 8 years ago by Anders Markvardsen

  • Status changed from new to accepted
  • Cc gesner.passos@… added

Steve/Richard showed yesterday that I had forgot to add the code for this to work in batch mode. Do this according to:

"We have agreed that the batch processor will now append “_front”, ”_rear” or ”_merged” onto output workspaces and file names to make batch processing much simpler."

Also note the sutblety:

"the SHIFT adjustment should only be made to a Sample run and not to a Can run, as it is a relative shift for the “sample-can” result for the front detector. The RESCALE adjustment needs to be applied to both Sample and Can runs. If the SHIFT and/or RESCALE are being fitted, this can be done on “sample-can” results for Front and Rear detectors, or just “sample” results if no can run is specified."

comment:4 Changed 8 years ago by Anders Markvardsen

Consisting logic for shift and rescale. re #5844

Changeset: 1b9bf07f33273cb71f1ed1d10ce300e4d7f860f9

comment:5 Changed 8 years ago by Anders Markvardsen

Added warnings when fit fail for shift and rescale. re #5844

It turns out that for LOQ reduced data can contain NaN entries. Code added to handle this case better

Changeset: 0dd19222c53863c487e3f5e195424b96ab342fc7

comment:6 Changed 8 years ago by Anders Markvardsen

  • Owner changed from Anders Markvardsen to Gesner Passos
  • Status changed from accepted to assigned

comment:7 Changed 8 years ago by Gesner Passos

  • Status changed from assigned to accepted

comment:8 Changed 8 years ago by Gesner Passos

The user King, warned that the reduction with the merging has these issues:

  • In single run mode the merging is not quite doing what it should, the main det data should not be rescaled or shifted, only the HAB/Front data should be;
  • The merging is not properly using the Can runs;
  • In batch mode no merged data is generated (the Q range of the output data is the same as that for the main det);
  • It should be possible to put DET/BOTH and DET/MERGED in a mask file to control a reduction, at the moment these have no effect.

comment:9 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detectors in batch mode.

isis_instrument.py:

  • add variable to allow selection of BOTH and MERGED

SANSBatchMode.py:

  • Extension BatchReduce function to accept a new input that informs which kind of reduction is expected and return an output of the scale and shift settings from the Fit algorithm.

ISISCommandInterface.py:

  • Extension of WavRangeReduction to provide new output (the scale and shift)
  • Refactoring of WavRangeReduction function for solve problem of having different paths from SingleMode and BatchMode
  • Propose a new function for the calculation of the merged function considering the CAN.

SANSRunWindow.cpp:

  • Correct bugs inside the loadUserFile function to correctly process the following entries: DET/[RESCALE|SHIFT]/FIT [Q1 Q2]
  • Extend loadUserFile to identify the options DET/[BOTH|MERGED|REAR|FRONT|HAB]
  • Correct the SingleMode Loading function to move correctly the detector - Causing different answer from batch and single mode.
  • Correct the python code related to SANSBatchMode and ISISCommandInterface
  • Update the shift and scale parameters at the enf of a Batch mode reduction.

re #5844

Changeset: 0ff0b453aea49e85fba4b390cfdad939b360f38d

comment:10 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detector

Fix the merging formula when no CAN is given (ISISCommandInterface) Implement the agrement to have _rear, _front, _merged suffix in batch mode (SANSBatchMode) re #5844

Changeset: ee41082310d595ad5b1a5aab8256993e54dd76cf

comment:11 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detector

Issue Warning message when the Fit Function returns a negative scale value. re #5844

Changeset: bd77e6a503606689600edc2cd3573e6fc012019e

comment:12 Changed 8 years ago by Gesner Passos

  • Status changed from accepted to verify
  • Resolution set to fixed

comment:13 Changed 8 years ago by Gesner Passos

  • Status changed from verify to reopened
  • Resolution fixed deleted

comment:14 Changed 8 years ago by Gesner Passos

Stephen reported some issues:

  • NameError: name fit_settings is not defined at line 2 in <Interface>
  • DET/MAIN is not being processed any more.

comment:15 Changed 8 years ago by Gesner Passos

SANS process DET/MAIN option

re #5844

Changeset: fa77edc28571c355cd135f5ac274758888f23da9

comment:16 Changed 8 years ago by Gesner Passos

Solve fit_settings not defined variable

When reduction fails, an extra warning message appears saying that there is no fit_settings variable available. This solves this problem.

re #5844

Changeset: 329c539122bd06d3aa4d7355437047ea4682b1de

comment:17 Changed 8 years ago by Gesner Passos

  • Status changed from reopened to accepted

comment:18 Changed 8 years ago by Gesner Passos

  • Status changed from accepted to verify
  • Resolution set to fixed

comment:19 Changed 8 years ago by Anders Markvardsen

  • Status changed from verify to verifying
  • Tester set to Anders Markvardsen

comment:20 Changed 8 years ago by Gesner Passos

Solve extra warnings that was confusing the user

Q1D2: correct the warning message to have WorkspaceIndex.

SANSRunWindow:

  • warning user that we will ignore the FLOOD file option when

the detector is the FRONT/HAB detector. A new ticket will be opened to deal with this properly.

  • Skip reloading the user file for every single reduction that caused

extra warnings to appear, confusing the user.

  • Accept DET/MERGE option to be passed as well as DET/MERGED.

ISISCommandInterface-> temporaly clear the FloodFile when reducting the FRONT/HAB detector.

re #5844

Changeset: 7ceaa78b4ad5d7f25cd44351e4e52e75429a1341

comment:21 Changed 8 years ago by Anders Markvardsen

  • Status changed from verifying to closed

Tested this with LOQ and SANS2D data and it looks good to me

comment:22 Changed 8 years ago by Anders Markvardsen

Consisting logic for shift and rescale. re #5844

Changeset: 1b9bf07f33273cb71f1ed1d10ce300e4d7f860f9

comment:23 Changed 8 years ago by Anders Markvardsen

Added warnings when fit fail for shift and rescale. re #5844

It turns out that for LOQ reduced data can contain NaN entries. Code added to handle this case better

Changeset: 0dd19222c53863c487e3f5e195424b96ab342fc7

comment:24 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detectors in batch mode.

isis_instrument.py:

  • add variable to allow selection of BOTH and MERGED

SANSBatchMode.py:

  • Extension BatchReduce function to accept a new input that informs which kind of reduction is expected and return an output of the scale and shift settings from the Fit algorithm.

ISISCommandInterface.py:

  • Extension of WavRangeReduction to provide new output (the scale and shift)
  • Refactoring of WavRangeReduction function for solve problem of having different paths from SingleMode and BatchMode
  • Propose a new function for the calculation of the merged function considering the CAN.

SANSRunWindow.cpp:

  • Correct bugs inside the loadUserFile function to correctly process the following entries: DET/[RESCALE|SHIFT]/FIT [Q1 Q2]
  • Extend loadUserFile to identify the options DET/[BOTH|MERGED|REAR|FRONT|HAB]
  • Correct the SingleMode Loading function to move correctly the detector - Causing different answer from batch and single mode.
  • Correct the python code related to SANSBatchMode and ISISCommandInterface
  • Update the shift and scale parameters at the enf of a Batch mode reduction.

re #5844

Changeset: 0ff0b453aea49e85fba4b390cfdad939b360f38d

comment:25 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detector

Fix the merging formula when no CAN is given (ISISCommandInterface) Implement the agrement to have _rear, _front, _merged suffix in batch mode (SANSBatchMode) re #5844

Changeset: ee41082310d595ad5b1a5aab8256993e54dd76cf

comment:26 Changed 8 years ago by Gesner Passos

SANS Rescale and Fitting and Merging Detector

Issue Warning message when the Fit Function returns a negative scale value. re #5844

Changeset: bd77e6a503606689600edc2cd3573e6fc012019e

comment:27 Changed 8 years ago by Gesner Passos

SANS process DET/MAIN option

re #5844

Changeset: fa77edc28571c355cd135f5ac274758888f23da9

comment:28 Changed 8 years ago by Gesner Passos

Solve fit_settings not defined variable

When reduction fails, an extra warning message appears saying that there is no fit_settings variable available. This solves this problem.

re #5844

Changeset: 329c539122bd06d3aa4d7355437047ea4682b1de

comment:29 Changed 8 years ago by Gesner Passos

Solve extra warnings that was confusing the user

Q1D2: correct the warning message to have WorkspaceIndex.

SANSRunWindow:

  • warning user that we will ignore the FLOOD file option when

the detector is the FRONT/HAB detector. A new ticket will be opened to deal with this properly.

  • Skip reloading the user file for every single reduction that caused

extra warnings to appear, confusing the user.

  • Accept DET/MERGE option to be passed as well as DET/MERGED.

ISISCommandInterface-> temporaly clear the FloodFile when reducting the FRONT/HAB detector.

re #5844

Changeset: 7ceaa78b4ad5d7f25cd44351e4e52e75429a1341

comment:30 Changed 8 years ago by Gesner Passos

Solve extra warnings that was confusing the user

Q1D2: correct the warning message to have WorkspaceIndex.

SANSRunWindow:

  • warning user that we will ignore the FLOOD file option when

the detector is the FRONT/HAB detector. A new ticket will be opened to deal with this properly.

  • Skip reloading the user file for every single reduction that caused

extra warnings to appear, confusing the user.

  • Accept DET/MERGE option to be passed as well as DET/MERGED.

ISISCommandInterface-> temporaly clear the FloodFile when reducting the FRONT/HAB detector.

re #5844

Changeset: 7ceaa78b4ad5d7f25cd44351e4e52e75429a1341

comment:31 Changed 5 years ago by Stuart Campbell

This ticket has been transferred to github issue 6690

Note: See TracTickets for help on using tickets.