DLBclass Classification Tool

This tool, written in Python, consists of a lightweight module in the form of a Jupyter notebook for visually aided classification. There are two major steps that need to be followed:

1. Setup - this includes cloning the repository, virtual environment setup, and package installation.

2. Running the Jupyter notebook: DLBclass_classify.ipynb

Setup

1. Users must have Python3 and git installed on their machine. If you already have both git and Python3 installed, move to the step #2 within the setup.

   
   

   Installing git

   open a terminal/command prompt and enter the command git --version. If your OS says something like "command not found", git needs to be installed, and so follow the instructions for your OS here:

   https://git-scm.com/book/en/v2/Getting-Started-Installing-Git

   
   

   Installing Python3

   Next, in the terminal enter python --version, and ensure that your version of Python is at least Python 3.7.x.

   If not, install the latest python version at:

   https://www.python.org/downloads/

   
   

2. Clone the repository to your local machine by entering the following commands:

   
   Mac / Linux users:

   cd ~/Desktop

   git clone [email protected]:getzlab/DLBclass-tool.git

   cd DLBclass-tool

   
   

   Windows users:

   chdir Desktop

   git clone [email protected]:getzlab/DLBclass-tool.git

   chdir DLBclass-tool

   
   

3. Setup your virtual environment. Enter the command:

   python -m venv ./venv

   
   

4. Activate your virtual environment. Enter the command:

   Mac / Linux users:

   source ./venv/bin/activate

   Windows users:

   .\venv\Scripts\activate.bat

   
   

5. Install required packages. This may take up to 5 minutes or so.

   pip install -r requirements.txt

Running the classification notebook

In the same command line session, enter the command:

jupyter notebook

once your browser opens the directory, click on the notebook named

DLBclass_classify.ipynb

Follow the instructions within the notebook to classify your samples.