This ticket describes the functionality of the ImportDataWizard to provide
an overview. Based on this information all pages should be layouted without
functionality.
Preconditions
[required] Patient selected via the IGlobalSelectionService
[optional] ISensor selected via the IGlobalSelectionService
[optional] Therapy via the IGlobalSelectionService
Function
Imports data into Medmon. There are different
types of data and sources which determine what happens after the import.
Data types
RAW: raw sensor data. Normally directly classified after import.
TRAINING: Data used to train the classification process. This data could be
RAW: musted be tagged afterwards
TRAINING: already tagged, directly imported. Normaly train classification process after import
Source types
Sensor: Raw data from a sensor. ISensorService is used to handle this data
CSV: Usually used to stored tagged data to easily import training data.
ARFF: Usually used to stored tagged data to easily import training data.
Import RAW
The data is stored via the IPatientService.
A "Select Classification Process" Page shows up to directly classify the data.
There should be an option "store results", which means the classified results
are stored too and linked to the TherapyResults entity. Thus the user has to
choose a Therapy entity.
Import TRAINING
There are two different possible pages after selecting "TRAINING" as data type:
source: Sensor || untagged CSV / ARFF
The data is raw and untagged. Tagging is a more complex task and should be
done in an own editor, maybe even in special perspective. This will be addressed
in another ticket.
source: CSV / ARFF (tagged)
CSV data are converted to ARFF an stored via the IPatientService.
WizardPages - Implementation hints
There should be a page for each clearly defined task. Maybe like
[Input]: Select Patient, Sensor and Type
on RAW input: [Classification]: Select DPU and Therapy
on TRAINING input (tagged): [Training]: Select DPU and update/new training profile
on TRAINING input (untagged): [TrainingPerspective]: directly open TrainingPerspective
Task
This ticket describes the functionality of the ImportDataWizard to provide an overview. Based on this information all pages should be layouted without functionality.
Preconditions
Function
Imports data into Medmon. There are different types of data and sources which determine what happens after the import.
Import RAW
The data is stored via the IPatientService.
A "Select Classification Process" Page shows up to directly classify the data. There should be an option "store results", which means the classified results are stored too and linked to the TherapyResults entity. Thus the user has to choose a Therapy entity.
Import TRAINING
There are two different possible pages after selecting "TRAINING" as data type:
source: Sensor || untagged CSV / ARFF
The data is raw and untagged. Tagging is a more complex task and should be done in an own editor, maybe even in special perspective. This will be addressed in another ticket.
source: CSV / ARFF (tagged)
CSV data are converted to ARFF an stored via the IPatientService.
WizardPages - Implementation hints
There should be a page for each clearly defined task. Maybe like
[Input]: Select Patient, Sensor and Type
Additional information
@see #38 @see de.lmu.ifi.dbs.medmon.medic.ui.wizard.QuickAnalyseWizard.java