-
Notifications
You must be signed in to change notification settings - Fork 22
Copy of datasets
The copy of datasets is a new feature that allow a user to duplicate dataset from a study to another. This can be used to share a subpart of a study to anyone instead of giving full access to all the data to someone. Datasets are not duplicated in PACS.
In order to keep the study's tree similar, the copy of datasets will also keep the examination, dataset acquisitions, centers and subjects linked to the selected datasets.
In order to copy datasets from a study to another, there are a few rules:
- You must be part of both studies (source and target) to copy datasets
- You must have admin rights on both studies
- You can't copy datasets that are copied from another study
- You can't copy a dataset to the study the dataset comes from
- You can't copy a dataset to a study that already contains the dataset
- You can't copy a processed dataset
The copy is creating new examinations, dataset acquisitions and datasets. If you want to delete a dataset, you must make sure no copy exist or a warning will be displayed. The same goes for examinations and dataset acquisitions.
If you select a processed dataset, it won't be copied but it won't throw an error either.
Create a new study, let's call it 'DestinationStudy'. Go to solr, filter and select the datasets that you want to copy, then click on the "copy to study" button
A dialog opens and asks you to select the study you want to copy those datasets in. Let's select 'DestinationStudy'.
Click "Copy" to start the copy. A new task will appear in the "Job" page, and once the progress hits 100%, the copy is done. You can compare both studies and see how similare they are
Once the copy is started, a new task will appear in shanoir's Job page. If many datasets are being copied, it can take some time (1000 datasets ~= 15mn). This task allows the user to see how the copy is doing, and, once done, show a report:
Since there are many condition to properly copy datasets, one could easily make a mistake (ex: select a processed dataset) and expect X datasets in the destination study. If there aren't X but Y datasets, the answer is in the task's report.