commandline_tutorial.md

December 10, 2021 · View on GitHub

Basic command-line instructions

Create/Save a project

myproject= shallowNew (then select a project name and place to save it)

shallowSave(myproject) (saves the project)

shallowLoad(path) (loads the project located in path; if no path is provided, launches a dialog box to select the project)

Import Data

myproject.addData allows to add new fields of view (FOV) to the project

ROIs

myproject.fov(myfovid).view allows to view the FOV myfovid and choose custom ROIs manually. For this, use the zoom & pan buttons and right click on the image

myproject.setPattern(myfovid,roiid,frameid) uses a given FOV/ROI/frame to create a pattern to be used for automated ROI detection

myproject.identifyROIs(myfovid,frameid) uses autocorrelation to identify all ROIs based on a given pattern in specific FOVs, using specific frame frameid

myproject.saveCroppedImages(myfovid,frameid) writes 4D volumes (w x h x ch x time) in a subfolder of the the main analysis folder

myproject.fov(myfovid).roi(myroiid).view allows to view the 4D volume in a GUI

myproject.fov(myfovid).roi(myroiid).traj(classistr) shows the "trajectory", ie the sequences of classification done according to a given classifier in a @classi object classistr= project.processing.classification(id).strid; Also works with ROIs that belong to @classi objects

Classification

Adding a new classification

myproject.addClassification create a new classifier ( @classi object) in this case, no ROI is imported in the classi object

myproject.processing.classification is an array that contains all classifiers created

myproject.addClassification(classiobject) will duplicate the @classi object user is asked whether to import ROIs included in the original classi object

myproject.addClassification(string) will import an existing classifier (string) from a repository %% not really implemented yet%%

Removing a classification

myproject.removeClassification(id) will remove a classifier specified by the index id.

Adding ROIs a to a classification

myproject.processing.classification(id).addROI(@classi object OR @fov object, optional: ROIs IDs) adds ROIs to the classification myproject.processing.classification(id). ROIs may come either from a @fov or from an exisiting @classi object This function will preserve training sets and reformat it if the number of classes are different ex: myproject.processing.classification(1).addROI(myproject.fov(1)) myproject.processing.classification(1).addROI(myproject.processing.classifiction(1).roi(1))

myproject.processing.classification(id).clearTraining to remove training data see function arguments for details

Using classifications

myproject.processing.classification(id).setClasses(classnames) allows the user to define and to reassign classes in the @classi object and dependencies (ROIs) number of classes can be extended or decreased

myproject.myproject.processing.classification(id).userTraining launches a GUI to set the ground truth using a specific ROI

myproject.fov(myfovid).roi(myroiid).fillTraining() Annotate classes of the training of the ROI, using the last annotated frame as template. Ex: [1 0 0 0 2 0 0 3] --> [1 1 1 1 2 2 2 3]

myproject.myproject.processing.classification(id).formatDataForTraining formats data to be used by the classifier; this must be done before launching the training procedure

myproject.myproject.processing.classification(id).setTrainingParam request parameters values for training the classifier; default values can be entered

myproject.processing.classification(id).trainClassifier

asks whether the trainingset needs to be formatted
asks whether training parameters must be updated
then trains the classifier

myproject.processing.classification(id).loadClassifier loads the classifier associated with @classi object in the workspace

myproject.processing.classification(id).validateTrainingData(optional: classifier) uses the classifier of the @classi object (optional: provide a classifier) to classify ROIs used to build the groundtruth in order to compare with classification results

myproject.classifyData(classifid,roilist,option) allows one to start the classification referred to as classifid on the roilist; You can provide the classifier as an option, so that it is not loaded each time you run the function

myproject.processing.classification(id).displayValidation loads a specific ROI along with the classification results and groundtruth if there are any This also provides basic statistics about the classification

myproject.processing.classification(id).stats compute and stores (as a txt file) the statistics associated with the classification and comparison to groundtruth

out=classifyData2(test.processing.classification(2),test.processing.classification(2).roi(1:2),'Classifier',classifier,'Parallel')

Extract RLS

rls=measureRLS2(theo.processing.classification(1),theo.processing.classification(1).strid)

measureRLS3(theoRLS.processing.classification(1),theoRLS.fov(1).roi(1:100))

statRLS(rls)

Extract signal from ROIs

myproject.fov(1).extractFluo(cf arguments below) outdated or myproject.processing.classification(id).extractFluo(cf arguments below) outdated Extract signal from the ROIs of the fov or classi object. Arguments: 'Method': 'maxPixels' computes the average of the kMaxPixels. // 'mean' 'Channels' 'Frames' 'Rois'

myproject.fov(1).detectFluoPattern('Channels',[4,5]) Arguments: *'Method': 'full' check .fluo.full.maxf // 'mean' checks the fluo.meanf *'Channels' *'Frames' *'Rois' *'fluoThreshold' *'frameThreshold' number of frames to be above fluoThreshold

Exporting movies or image sequences

Coming soon/In construction myproject.fov.export('Frames',1:5,'Framerate',10,'FontSize',96,'Levels',[4000 14000],'DrawROIs',[],'Drift')

myproject.processing.classification(3).export('Mosaic',1:9,'Name','test','Training','Results','Levels',[6000 20000],'Framerate',10,'Title','fob1','RLS')

myproject.fov(1).roi(1).export('Frames',[1:10:150],'Sequence',3,'Levels',{[5000 30000]},'Background',[1 1 1],'Text',[0 0 0],'Training','Results','Classification',rls.processing.classification(1),'RLS') export sequence of Frames

Make independant classifications

list = listRepositoryClassi; will create a .txt file in the default matlab folder, containing the indicated path for a folder in which you will put all your desired independant classifications. In this folder, you will be able to create new classi, export existing classi into shallow projects, update existing classi, import classi into it, and add/update ROIs to existing classi. to be continued...