Installation

Step by step instructions are below.

If you need help, please join our email list at hvc-users@google-groups.com and ask a new question.

Note that the $ and > prompts below are just to indicate that you’re in the command line, you don’t have to type them. If the command is the same on Mac/Linux/Windows then only the $ prompt is shown.

Easy install for general users

using the Anaconda distribution and the conda package manager

The following should work on Linux, Mac, and Windows.

1. Install the Anaconda distribution for your operating system: https://www.anaconda.com/download/

2. Add the conda forge channel which contains some of the libraries necessary for hybrid-vocal-classifier to run.

$ conda config --add channels conda-forge

For more about conda forge, see: https://conda-forge.org

If you have previously added this channel, you do not need to do so again.

You can check whether it’s added by running the following command:

$ conda config --show

If the channel is added, you should see something like the following lines in the config output:

channels:
- conda-forge

3. Create a conda environment.

$ conda create -n hvc python=3.5

Currently the environment must use Python 3.5.

The environment allows you have to install the libraries necessary for hybrid-vocal-classifier to run.

For a more in-depth explanation see https://conda.io/docs/user-guide/concepts.html#conda-environments.

4. Activate the environment

Mac/Linux:

$ source activate hvc

Windows:

> activate hvc

After you activate the environment, its name will appear in parentheses before the terminal prompt.

(hvc)$

5. Install hybrid-vocal-classifier into the environment

(hvc)$ conda install -c nickledave hybrid-vocal-classifier

6. Test whether the install worked.

(hvc)$ python

>>> import hvc

If the above line executes without any module not found error, you have successfully installed hybrid-vocal-classifier.

Install of bleeding-edge / development version

using the Anaconda distribution and the conda package manager

1. Install the Anaconda distribution for your operating system: https://www.anaconda.com/download/

2. Use `conda to create an environment for the bleeding-edge version

$ conda create --name hvc-bleeding-edge python=3.5 numpy scipy scikit-learn matplotlib pyyaml keras tensorflow

conda will ask you if you want to install these packages and their dependencies, say [y]es.

You may need to add the conda forge channel in order to find binaries of these packages that will work on your

operating system. To add the channel, you execute the following:

$ conda config --add channels conda-forge

More about conda forge here: https://conda-forge.org/

3. git clone the repository

(hvc-bleeding-edge) $ git clone https://github.com/NickleDave/hybrid-vocal-classifier.git

4. Activate the environment so you can work with it.

$ source activate hvc-bleeding-edge

5. use pip to install the hvc code into the conda environment, using the “-e” flag (for “editable”):

(hvc-bleeding-edge) pip install -e hybrid-vocal-classifier

To do this, you need to be in the parent directory just above the hybrid-vocal-classifier directory

(which you probably already are if you just executed the git clone command)

so that pip can find the necessary setup.py file for the install.

Now whenever you want to get the most up-to-date version you can execute

(hvc-bleeding-edge) $ git pull

and as long as you haven’t made any changes to the code base,

git should just pull new changes in from the remote and merge them with the old version.

You probably also want to install Jupyter and iPython in the bleeding-edge environment.

(hvc-bleeding-edge) $ conda install ipython jupyter

Beware: confusingly, you can start iPython and Jupyter from the command line

even if they are not installed in your environment,

but you will be running the versions in the root conda environment,

and so iPython and Jupyter won’t know that hvc et al. are installed.

You should now be able to start iPython or a Jupyter notebook and import hvc to work with it.