Load a BIDS Dataset

Load a BIDS Dataset#

Before you can query or modify your data, you need to make your dataset available to ancpBIDS. This involves two steps: fetching a dataset and creating a layout to represent its structure.

1. Fetch an existing BIDS dataset#

ancpBIDS is able to load specific files from your PC. With ´fetch_dataset()´, ancpBIDS will be able to load it even if you haven’t unzip your dataset.

from ancpbids import utils
dataset_path = utils.fetch_dataset('ds005')

from ancpbids import utils
dataset_path = utils.fetch_dataset('ds003483')

The output variable, dataset_path, will contain the local path to your dataset.

Also, within your PC folder

home/user/.ancp-bids/datasets

you will find the zip file (f.e. ds005-testdata.zip) and the content extracted (f.e. ds005). If you run the code again, it won’t create unnecessary copies.

Don’t have a Dataset?

In case you don’t have a BIDS compliant dataset, you can download a test dataset from our github. We offer you two types: a MEG dataset (ds003483) and a MRI dataset (ds005). If you have your own BIDS dataset, feel free to use yours instead. These datasets are only meant to learn how to use ancpBIDS, and are not expected to be used in any kind of research.

2. Creating a Layout#

Once you have your dataset available locally on your PC, we can use ´BIDSLayout()´ to create the in-memory graph of it, from where you can easily retrieve information. Both the dataset and the schema are accessible through the layout object.

from ancpbids.pybids_compat import BIDSLayout
layout = BIDSLayout(dataset_path)

In-memory graph?

If you want to learn more how ancpBIDS uses the BIDS specification to build the in-memory graph representation (and what exactly is a in-memory graph representation), follow this link.

Load a dataset#

Thanks to BIDSLayout your dataset it’s already loaded within the layout variable.

Alternatively, you may also use the function ´load_dataset()´ along with the dataset_path you fetched before to load the dataset.

from ancpbids import load_dataset
dataset = load_dataset(dataset_path)
#print(dataset)
#{'name': 'ds003483'}

from ancpbids import load_dataset, DatasetOptins
dataset = load_dataset(dataset_path, DatasetOptions(ignore=False, infer_artifact_datatype=True))
#print(dataset)
#{'name': 'ds003483'}

Next section#

Now that your dataset is loaded and mapped into a layout, you’re ready to start extracting useful information from it. In the next sections, we’ll show how to:

Contents