Link Search Menu Expand Document

Creating a new data model


All of the data models used in MNE Analyze are subclassed from the same base, AbstactModel (applications/mne_analyze/libs/anShared/Model/abstactmodel.h).

Subclassing AbstactModel

Abstract model is itself a subclass of an abstract class, QAbstactItemModel. To correctly sub-class it, all of the pure virtual functions from QAbstractItemModel and AbstactModel must be implemented. For more information on the QAbstactItemModel functions, see the Qt documentation.

The following are the functions that need to be implemented:


Returns the type of the model. These types are declared in applications/mne_analyze/libs/anShared/Utils/types.h, and when creating a new model, a new item needs to be added to enum MODEL_TYPE.


Returns data stored in model based on the input model index and role.


Returns the flags for the model index passed to the function.


Return an index for the data based on the input row and column.


Return a model index to the parent of this model. IF there is no parent, return an empty QModelIndex


Return an appropriate value depending on how you are organizing or representing your data. The FiffRawViewModel, for example, returns the amount of channels for which it has data.


Return an appropriate value depending on how you are organizing or representing your data. The Annotation Model, for example, returns the number of columns of data it has to display.

Loading/Adding to Data Loader and Data Manager

If your data model will be loading data from a file, make sure that file type is included in DataLoader::loadFilePath() so the user can load in that file type. To save to a file, implement the function saveToFile() from the AbstractModel, and ensure that your file type is included in DataLoader::onSaveFilePressed().

To add your model to the data manager, add it to either addModel or loadModel in AnalyzeData.h, depending on whether you will be loading data from inside the application or from a file. You need to define the type of data your model corresponds to, bidsviewmodel.h for possible types, or add a new one.