NSXTool issueshttps://code.ill.fr/scientific-software/nsxtool/-/issues2018-12-11T17:23:26+01:00https://code.ill.fr/scientific-software/nsxtool/-/issues/314replace NSXTest by standard test framework2018-12-11T17:23:26+01:00Joachim Wuttkereplace NSXTest by standard test frameworkAll this functionality should exist out of the box. Let's try catch2.All this functionality should exist out of the box. Let's try catch2.Code cleanuphttps://code.ill.fr/scientific-software/nsxtool/-/issues/313replace libyaml-cpp by libyaml2018-12-08T11:13:56+01:00Joachim Wuttkereplace libyaml-cpp by libyamllibyaml-cpp is unstable, bogus, slow, and introduces an unstable boost dependence. The C library libyaml is the safer choice.libyaml-cpp is unstable, bogus, slow, and introduces an unstable boost dependence. The C library libyaml is the safer choice.Removing boost dependencyhttps://code.ill.fr/scientific-software/nsxtool/-/issues/312replace #pragma by include guards2018-12-11T17:23:21+01:00Joachim Wuttkereplace #pragma by include guards#pragma may seem more elegant than conventional include guards, but it is not standard.#pragma may seem more elegant than conventional include guards, but it is not standard.Code cleanuphttps://code.ill.fr/scientific-software/nsxtool/-/issues/300plot statistics against d instead of shell id2018-11-16T12:47:00+01:00eric pellegriniplot statistics against d instead of shell idhttps://code.ill.fr/scientific-software/nsxtool/-/issues/299update peak table view when normalizing to monitor2018-11-27T21:21:58+01:00eric pellegriniupdate peak table view when normalizing to monitoreric pellegrinieric pellegrinihttps://code.ill.fr/scientific-software/nsxtool/-/issues/274compute I/sigma of forbidden peaks in the space group dialog2019-01-02T16:15:59+01:00Jonathan Fishercompute I/sigma of forbidden peaks in the space group dialogTo help users select the correct space group, we should compute the mean I/sigma of the forbidden peaks in the space group dialog and list it alongside the compatibility percentage.To help users select the correct space group, we should compute the mean I/sigma of the forbidden peaks in the space group dialog and list it alongside the compatibility percentage.https://code.ill.fr/scientific-software/nsxtool/-/issues/261save session model to YAML format2019-01-02T16:20:09+01:00eric pellegrinisave session model to YAML formatThis work has been started a while ago using JSON format. The fact that nsxtool is now relying on YAML for databases and instruments storage/serialization make that we should switch to this format to going on with that work.This work has been started a while ago using JSON format. The fact that nsxtool is now relying on YAML for databases and instruments storage/serialization make that we should switch to this format to going on with that work.https://code.ill.fr/scientific-software/nsxtool/-/issues/251add detector efficiency to instrument file and correct intensities2019-01-02T16:20:26+01:00Jonathan Fisheradd detector efficiency to instrument file and correct intensitiesIntensities should be scaled as a function of angle of incidence and detector efficiency (see e.g. Denzo X files).
For good efficiency (e.g. 99%) this is likely negligible, but for even 90% this becomes significant.Intensities should be scaled as a function of angle of incidence and detector efficiency (see e.g. Denzo X files).
For good efficiency (e.g. 99%) this is likely negligible, but for even 90% this becomes significant.https://code.ill.fr/scientific-software/nsxtool/-/issues/2473d Peak integrators should compute rocking curve2018-06-29T17:07:29+02:00Jonathan Fisher3d Peak integrators should compute rocking curveeric pellegrinieric pellegrinihttps://code.ill.fr/scientific-software/nsxtool/-/issues/246compute uncertainty from fit profile2019-01-02T16:21:20+01:00Jonathan Fishercompute uncertainty from fit profileThe fit profile itself has uncertainty due to the way it is calculated, but this is not used in the estimated uncertainty of fitted intensities.The fit profile itself has uncertainty due to the way it is calculated, but this is not used in the estimated uncertainty of fitted intensities.Jonathan FisherJonathan Fisherhttps://code.ill.fr/scientific-software/nsxtool/-/issues/245merge FitProfile and IntegratedProfile into a single n-d Profile class2019-01-02T16:20:51+01:00Jonathan Fishermerge FitProfile and IntegratedProfile into a single n-d Profile classNote: should store vector of Intensity, not just double, so that we can compute errors.
Use the sampling method described in the Kabsch XDS paper.Note: should store vector of Intensity, not just double, so that we can compute errors.
Use the sampling method described in the Kabsch XDS paper.Jonathan FisherJonathan Fisherhttps://code.ill.fr/scientific-software/nsxtool/-/issues/243CC factor statistics2018-05-14T14:26:10+02:00Jonathan FisherCC factor statisticsThe CC factors are random variables, which depend on a splitting of the data into two halves. Since it is cheap to compute, we should compute CC by taking an average over a number of random splittings.The CC factors are random variables, which depend on a splitting of the data into two halves. Since it is cheap to compute, we should compute CC by taking an average over a number of random splittings.https://code.ill.fr/scientific-software/nsxtool/-/issues/241autoindex: provide list of UC symmetry constraints2019-01-02T16:21:45+01:00Jonathan Fisherautoindex: provide list of UC symmetry constraintsAutoindex should first fit _without_ symmetry constraints, and then for each Bravais type (Niggli reduction) generate the best fit to the reference cell and output percentage of peaks still indexed.Autoindex should first fit _without_ symmetry constraints, and then for each Bravais type (Niggli reduction) generate the best fit to the reference cell and output percentage of peaks still indexed.https://code.ill.fr/scientific-software/nsxtool/-/issues/240store reference peaks library in experiment tree2019-01-02T16:22:41+01:00Jonathan Fisherstore reference peaks library in experiment treeInstead of a single "peaks" item we should have (at least) two: reference peaks and predicted peaksInstead of a single "peaks" item we should have (at least) two: reference peaks and predicted peaksJonathan FisherJonathan Fisherhttps://code.ill.fr/scientific-software/nsxtool/-/issues/239tasks for shape library + dialog2019-01-02T16:22:21+01:00Jonathan Fishertasks for shape library + dialog* "Fit peak profile" in GUI should be "Build shape library"
* Peak and background scale parameters should be saved as part of library and used for peak prediction and integration
* Peak prediction and integration dialogs should retrieve ...* "Fit peak profile" in GUI should be "Build shape library"
* Peak and background scale parameters should be saved as part of library and used for peak prediction and integration
* Peak prediction and integration dialogs should retrieve peak and background scale from shape library
* Dialog should also give 1d predicted profile
* Number of bins for 1d profile should be exposed in dialog
* Dialog should expose option to switch between detector coords and Kabsch coordsJonathan FisherJonathan Fisherhttps://code.ill.fr/scientific-software/nsxtool/-/issues/238q values should be weighted during UB refinement2018-05-04T17:25:23+02:00Jonathan Fisherq values should be weighted during UB refinementCurrently all q values are treated equally during refinement, which is not correct. The should probably be weighted by covariance matrices and possibly also I/sigma.Currently all q values are treated equally during refinement, which is not correct. The should probably be weighted by covariance matrices and possibly also I/sigma.Jonathan FisherJonathan Fisherhttps://code.ill.fr/scientific-software/nsxtool/-/issues/237proper treatment of pixel saturation and streaking2019-01-02T16:24:28+01:00Jonathan Fisherproper treatment of pixel saturation and streakingOutstanding issue which definitely affects D19 and occasionally affects Biodiff. Proper treatment should definitely improve R factors at low resolution.Outstanding issue which definitely affects D19 and occasionally affects Biodiff. Proper treatment should definitely improve R factors at low resolution.https://code.ill.fr/scientific-software/nsxtool/-/issues/233implement strategy design pattern for integration methods2018-04-26T21:07:40+02:00eric pellegriniimplement strategy design pattern for integration methodsInstead of having `IPeakIntegrator` interface and all of the integrators deriving from ti, we could have concrete `PeakIntegrator` that takes `IIntegrationStrategy` as argument in its `integrate` method. The strategy pattern may be suite...Instead of having `IPeakIntegrator` interface and all of the integrators deriving from ti, we could have concrete `PeakIntegrator` that takes `IIntegrationStrategy` as argument in its `integrate` method. The strategy pattern may be suited for this. Moreover, the integration strategies should be stored in a factory so that we can query the integration modes and their parameters in the GUIeric pellegrinieric pellegrinihttps://code.ill.fr/scientific-software/nsxtool/-/issues/225add min and max options to DetectorScene2018-04-26T10:25:31+02:00Jonathan Fisheradd min and max options to DetectorSceneE.g. in matplotlib's imshow, the image is transformed as
```
intensity -> (intensity - min) / (max-min)
```
where max and min are the maximum and minimum intensities _of the visible frame_. This is very convenient for visualizing weak pe...E.g. in matplotlib's imshow, the image is transformed as
```
intensity -> (intensity - min) / (max-min)
```
where max and min are the maximum and minimum intensities _of the visible frame_. This is very convenient for visualizing weak peaks in protein crystal data.
In contrast, in DetectorScene we scale the image as `intensity -> intensity / max`, where max is a user-defined value. Max being user-defined is not so bad, but without also subtracting a minimum value, it makes it very difficult to visualize weak peaks.eric pellegrinieric pellegrinihttps://code.ill.fr/scientific-software/nsxtool/-/issues/221support importing tiff in raw data importer2018-06-13T15:17:41+02:00Jonathan Fishersupport importing tiff in raw data importereric pellegrinieric pellegrini