Instructions to run the tutorial notebooks for the [Allen Brain Observatory dataset] (http://observatory.brain-map.org/visualcoding/sdk/index).
Install Python from https://www.continuum.io/downloads.
Install git https://git-scm.com/book/en/v2/Getting-Started-Installing-Git
(On Windows use git bash as a shell to run the following commands)
Downloading all the files.
git clone https://github.com/alleninstitute/AllenSDK/
cd AllenSDK/doc_template/examples/nb
The AllenSDK is Python 2 compatible for now. So if you have Python 2 you can install it directly.
pip install allensdk
jupyter notebook
You can browse through the notebooks and run them cell by cell.
If you have Python 3 you can run the notebooks within a virtual environment for Python 2:
conda create --name py2allen python=2.7 anaconda
activate py2allen (on Windows)
source activate py2allen (on Linux, Mac)
pip install allensdk
jupyter notebook
Once you are done with work you can deactivate the environment:
deactivate (on Windows)
source activate (on Linux, Mac)
- get the notebooks
git clone https://github.com/valentina-s/AllenSDK
This creates a directory AllenSDK in the current folder. The notebooks are in the subfolder doc_template/examples/nb:
cd AllenSDK/doc_template/examples/nb
ls
- start the notebooks
jupyter notebook
The brain_observatory.ipynb notebook is a good place to start.
When running the boc.get_ophys_experiments function several files are created in the boc/ophys_experiment_data/ each with extension .niw, i.e. Neurodata Without Borders format.
These files can be treated as .hdf5 format files and can be read with any HDF5 reader (in Python with the h5py package). They have nested structure so one need to go deeper to get the data.