Installation

GUIMOV has differents way to be installed. According your utilisation you would use pip, docker or from source for installed it. After the installation look at datasets requirements to add your own datasets, then tutorials to learn how use GUIMOV.

pip

GUIMOV is available as a package. It is not already on PyPI but you could install it easily with this code.

>>> pip install git+https://github.com/PeyricT/guimov.git

using Command line

>>> guimov_launch --datasets Path/to/your/datasets/file --logs Path/to/your/logs/file

If you want explore a single dataset, prefere using the following command

>>> guimov_launch --singledataset Path/to/your/dataset

To start a quick demo use the following command

>>> guimov_launch -s demo

Using Python

then you would be able to use GUIMOV as standar python package. In order to use your own datasets you can use guimov.settings, the sttings must be set before launch the app, therefore default settings will be used and paths must not be found.

>>> import guimov
>>> guimov.settings.datasets_path = 'Path/to/your/datasets/file'
>>> guimov.settings.logs_path = 'Path/to/your/logs/file'
>>> guimov.start()

If you want explore a single dataset, prefere using the following command

>>> guimov.start(dataset='Path/to/your/dataset')

Docker

Using Docker is the easyest way to install GUIMOV, you only have to download the docker image and load it. With docker you can’t move the path to access datasets. However you can mount your directory inside the container.

>>> docker run -d -p 8050:8050 -v /Users/Username/your_datasets:/app/datasets -v /Users/Username/your_logs:/app/logs guimovapp

From source

GUIMOV provide an requirements.txt file, in order to quickly install all dependencies.

>>> git clone https://github.com/PeyricT/guimov.git
>>> cd guimov
>>> pip install -r requirements.txt

Then you can start the interface with the following command :

>>> python guimov/__main__.py

Here the best way to add your directory is adding them in datasets directory.

Servers mode

GUIMOV could be install on server same way as above.