IntroductionΒΆ
CAJAL is a toolbox to interface with the NeuroData image and annotation spatial databases. This provides an easy method for users to create and interact with data across a variety of use cases. For client applications, users will generally follow these steps:
- Server setup
- Upload Image Data (if not using an existing data set)
- Create and Upload Annotation Data
- Update Annotations (if needed)
- Delete Annotations (if needed)
- Query (Read) Annotations
For most of these usecases, we have created examples, functions, and loni modules to get you started.
Although documentation exists for the API at a method and object level, we have provided functions and wrappers to allow most of the database interface to happen via simple code reuse and drag-and-drop modules.
CAJAL Leverages RAMON, our flexible data standard to facilitate data exchange and processing. More details are available here: http://docs.neurodata.io/nddocs/ramon.html
This website documents the major functionality of CAJAL; for more information, please read through the example scripts and unit tests. CAJAL was written to be very verbose; in MATLAB, tab completing is very helpful for enumerations, setting fields, and exploring available options. Reading classes (e.g. edit RAMONSynapse) may also be of use.