GUI usage

bmiptools is equipped also with a basic graphical interface. In order to use it, one need to have a python terminal installed on its PC. Here the one provided by Anaconda is used.

Run the GUI

To run the bmiptool GUI it open the Anaconda Prompt and activate the bmiptools environment. At this point one has to execute the run_gui module of bmiptools with the python interpreter using the -m option. More precisely, assuming bmiptools has been installed in the environment called bmiptools_env, the activation of the environment and the launch of the GUI can be done with the two following line

$ conda activate bmiptools_env
$ (bmiptools_env) python -m bmiptools.run_gui

The short clip below show how this can be done one a Windows 10 compute with anaconda already installed.

Load a stack

The clip below show how a stack of tiff images can be loaded from a folder. The loading operation terminates when "Stack loaded!" is printed on the terminal.

Create and initialize a pipeline

The creation and initialization of a pipeline with the bmiptool GUI is showed in the clip below. By clicking on the create and initialize pipeline, a windows where the pipeline can be created by adding the various plugins. In this window one has also to specify the working folder of the pipeline. To initialize each plugin click on the setting button and, once the plugin configuration windows opens, click on ok, after modifying the parameters if needed. When this operation is fully executed, "Pipeline created and initialized!" appears on the terminal.

Attention

In order to add and correctly initialize the plugins in the pipeline, always click on the setting setting and then click on ok, possibly after modifying the parameters.

Load a pipeline template

The creation and initialization of a pipeline can happens also ina different way: via a pipeline template. This is a json file containing all the pipeline parameters (see here to get more information about its structure). At the end of this operation "Pipeline template loaded!" is printed on the terminal.

Apply pipeline and previews

To apply the pipeline crated to the stack (running also the plugins optimizations if available) simply click on the button Apply pipeline. If a preview of the final result is needed, before to click on ‘Apply pipeline’ one needs to configure the preview. This can be done by clicking Preview setting, specify the preview setting and then apply the pipeline. Once that the pipeline has been applied on the stack, Pipeline applied! is printed on the terminal.

Save results

Once that pipeline has been applied to the stack, one can do two things: save the resulting stack and save the trained pipeline. To save stack simply click on Save stack: when the stack is saved in the selected folder "Stack saved!" is printed on the terminal. To save the pipeline click on Save pipeline: a .dill file will be produced in the pipeline folder selected, while the pipeline json will be updated with the parameters found during the optimization routine (if any). The clip below shows both these operations.

Load an existing pipeline

Once that a pipeline has been saved, a ‘.dill’ file is created in the pipeline folder (plus eventually a folder called ‘undillable’, see here for more information). This file is the one containing all the pipeline parameters (possibly determined by the optimization procedure) and can be read by bmiptools to reproduce exactly the same pipeline later or on a different computer. the clip below show how to load a pipeline with the bmiptoos GUI. When the loading is terminated, "Pipeline loaded!" appears in the terminal window.

At this point, one can apply the pipeline simply by clicking on Apply pipeline button. In this case no optimization routines are executed, and the (already trained) plugins are simply applied to the stack.