# Install Guide¶

## OS Support¶

Nodeworks is a pure python library. As a result, all operating systems that python and the dependencies can be installed on will run nodeworks.

## Installation¶

There are two methods for installing Nodeworks: 1. Use conda, which will install the required dependencies; or 2. Use poetry, which will create a virtual environment; or 2. From source, which will require you to manually install the dependencies

Note

If you are install Nodeworks as a plugin to MFiX, please see the MFiX-plugin specific documentation.

### Conda install¶

The easiest and recommended way to install nodeworks and the dependencies is through the Anaconda Distribution. Download and install Miniconda or Anaconda.

Once Miniconda or Anaconda are installed:

• Copy the conda install command. The conda command will look similar to this, but with your own individual username and hexadecimal token (please do not share these unique tokens):

conda install nodeworks -c mfix -c https://mfix.netl.doe.gov/s3/HEX/HEX/conda/dist

• Paste the conda install command into your Linux or Mac terminal, or into your Anaconda Prompt on Windows and Execute the install command.

Note

On Windows, to open the Anaconda Prompt, go to the start menu and search For Anaconda Prompt. The Anaconda Prompt sets up the correct environment, adding conda and other commands.

The installation will download other dependencies and Nodeworks. The process will take a few minutes to complete. Once completed, the application can be started by entering the following command into your Linux or Mac terminal, or into your Anaconda Prompt on Windows:

nodeworks


Note

On Windows, a desktop icon might be created, depending on user permissions.

### Using Poetry¶

Poetry is a tool for managing dependencies. To use poetry, first install poetry if you do not already have it. Then follow the following steps:

• Download the tarball, extract, and go into the top level source directory:

tar xf nodeworks-*.tar.gz
cd nodeworks

• install nodeworks with:

poetry install

• Run Nodeworks with:

poetry run nodeworks


### From source¶

If you already have a python environment setup without using Miniconda or Anaconda, nodeworks can be installed from source:

• Download the tarball, extract, and go into the top level source directory:

tar xf nodeworks-*.tar.gz
cd nodeworks

• install nodeworks with:

python setup.py install


Please see the dependencies listed below. Once completed and the dependencies are met, the application can be started by using the following command:

nodeworks


Note

To build the local documentation, the following libraries are required: - sphinx - sphinx_rtd_theme

## Required Dependencies¶

If the conda install command was used as recommended above, these dependencies have already been satisfied. If you installed from source or are missing nodes, the following libraries are required:

• A Qt wrapper: PyQt4, PyQt5, or PySide

• qtpy

qtpy can easily be installed via:

pip install qtpy


These requirements can be easily met by using Anaconda.

## Node Dependencies¶

Nodes can have other dependencies, based on their functionality. These dependencies are soft, meaning that they will not prevent nodeworks from, loading. They will only prevent nodes that depend on these dependencies from loading.

If using Miniconda or Anaconda, these dependencies can be easily installed with the following conda command:

conda install <dependency>


### Surrogate modeling and analysis toolset¶

Note

If you installed Nodeworks using the conda command (i.e. not from source), these dependencies have already been installed. You may skip this section.

For the Surrogate modeling and analysis toolset, the following libraries are required:

• numpy

• matplotlib

• scikit-learn

• scipy

• SALib

• py-earth

most of the libraries are in conda and can easily be installed with:

conda install numpy matplotlib scikit-learn scipy


py-earth is an add-on to scikit-learn, providing the MARS response surface model. It can be installed with conda:

conda install -c conda-forge sklearn-contrib-py-earth


salib is used to perform sensitivity analysis. It can also be installed with conda:

conda install -c conda-forge salib


### image processing¶

For image processing nodes, these libraries are used:

• opencv

• scikit-image

• imageio

• numpy

• scipy

and can easily be installed with conda:

conda install opencv scikit-image imageio numpy scipy


### wavelets¶

For wavelet analysis, the following library is used:

• pywavelets

and can easily be installed with conda:

conda install pywavelets


## Update¶

To update nodeworks, follow the above install instructions

## Uninstall¶

If installed with conda, uninstall by running the following command:

conda uninstall nodeworks


If installed from source, uninstall by running:

pip uninstall nodeworks