wrf-python ¶

For more information, see: https://conda.io/miniconda.html

Note

What is Miniconda?

If you have used the Anaconda distribution for Python before, then you will be familiar with Miniconda. The Anaconda Python distribution includes numerous scientific packages out of the box, which can be difficult for users to build and install. More importantly, Anaconda includes the conda package manager.

The conda package manager is a utility (similar to yum or apt-get) that installs packages from a repository of pre-compiled Python packages. These repositories are called channels. Conda makes it easy for Python users to install and uninstall packages, and also can be used to create isolated Python environments (more on that later).

Miniconda is a bare bones implementation of Anaconda and only includes the conda package manager. Since we are going to use the conda-forge channel to install our scientific packages, Miniconda avoids any complications between packages provided by Anaconda and conda-forge.

Step 3: Install Miniconda¶

Windows:

Browse to the directory where you downloaded Miniconda3-latest-Windows-x86_64.exe.

Double click on Miniconda3-latest-Windows-x86_64.exe.

Follow the instructions.

Restart your command terminal.

Mac and Linux:

For Mac and Linux, the installer is a bash script.
Using a terminal, you need to execute the bash shell script that you downloaded by doing:
bash /path/to/Miniconda3-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda3-latest-Linux-x86_64.sh [Linux]
Follow the instructions.
At the end of the installation, it will ask if you “wish the installer to initialize Miniconda3”. If you are unsure what to do, you should say “yes”. If you say “no”, we’re going to assume you know what you are doing.

If you said “yes”, then once you restart your shell, the “base” conda environment will be activated by default and the miniconda3 Python will be found instead of the system Python when you type the “python” command. If you want to undo this later, then you can run the following “conda” command:
conda config --set auto_activate_base false
Note that this will not affect your ability to run “conda” commands, it will just prevent the miniconda3 Python from overriding any existing Python environments you may already have on your machine.
Restart your command terminal.
[Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then you need to activate the bash shell by typing the following in to your command terminal:
bash
Verify that your system is using the correct Python interpreter by typing the following in to your command terminal:
which python
You should see the path to your miniconda installation. If not, see the note below.

Note

If you have already installed another Python distribution, like Enthought Canopy, you will need to comment out any PATH entries for that distribution in your .bashrc or .bash_profile. Otherwise, your shell environment may pick the wrong Python installation.

If bash is not your default shell type, and the PATH variable has been set in .bash_profile by the miniconda installer, try executing “bash -l” instead of the “bash” command in step 5.

Step 4: Set Up the Conda Environment¶

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when working with conda.

The name of our conda environment for this tutorial is: tutorial_2019.

Follow the instructions below to create the tutorial_2019 environment.

Open a command terminal if you haven’t done so.
[Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not your current shell, type:
bash
Add the conda-forge channel to your conda package manager.

Type or copy the command below in to your command terminal. You should run this command even if you have already done it in the past. This will ensure that conda-forge is set as the highest priority channel.
conda config --add channels conda-forge
Note

Conda-forge is a community driven collection of packages that are continually tested to ensure compatibility. We highly recommend using conda-forge when working with conda. See https://conda-forge.github.io/ for more details on this excellent project.
Create the conda environment for the tutorial.

Type or copy this command in to your command terminal:
conda create -n tutorial_2019 python=3.7 matplotlib cartopy netcdf4 jupyter git ffmpeg wrf-python
Type “y” when prompted. It will take several minutes to install everything.

This command creates an isolated Python environment named tutorial_2019, and installs the python interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.
Note

When the installation completes, your command terminal might post a message similar to:
If this is your first install of dbus, automatically load on login with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda3/envs/tutorial_test/org.freedesktop.dbus-session.plist ~/Library/LaunchAgents/
launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-session.plist
This is indicating that the dbus package can be set up to automatically load on login. You can either ignore this message or type in the commands as indicated on your command terminal. The tutorial should work fine in either case.
Activate the conda environment.

To activate the tutorial_2019 Python environment, type the following in to the command terminal:

For Linux and Mac (using bash):
conda activate tutorial_2019
For Windows:
activate tutorial_2019
You should see (tutorial_2019) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:

For Linux and Mac:
conda deactivate
For Windows:
deactivate tutorial_2019

Step 5: Download the Student Workbook¶

The student workbook for the tutorial is available on GitHub. The tutorial_2019 conda environment includes the git application needed to download the repository.

These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another directory, we’re going to assume you know how to do this yourself.

To download the student workbook, follow these instructions:

Activate the tutorial_2019 conda environment following the instructions in the previous step (conda activate tutorial_2019 or activate tutorial_2019).
Change your working directory to the home directory by typing the following command in to the command terminal:

For Linux and Mac:
cd ~
For Windows:
cd %HOMEPATH%
Download the git repository for the tutorial by typing the following in to the command terminal:
git clone --recursive https://github.com/NCAR/wrf_python_tutorial.git
There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes, type the following in to the command terminal:

For Linux and Mac:
conda activate tutorial_2019

cd ~/wrf_python_tutorial/wrf_workshop_2019

git pull
For Windows:
activate tutorial_2019

cd %HOMEPATH%\wrf_python_tutorial\wrf_workshop_2019

git pull
Note

If you try the “git pull” command and it returns an error indicating that you have made changes to the workbook, this is probably because you ran the workbook and it contains the cell output. To fix this, first do a checkout of the workbook, then do the pull.
git checkout -- .
git pull

Step 6: Verify Your Environment¶

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see some warnings for matplotlib or xarray, but you can safely ignore these).

Activate the tutorial_2019 conda environment if it isn’t already active (see instructions above).
Open a python terminal by typing the following in to the command terminal:
python
Now type the following in to the Python interpreter:
>>> import netCDF4
>>> import matplotlib
>>> import xarray
>>> import wrf
You can exit the Python interpreter using CTRL + D

Step 7: Obtain WRF Output Files¶

The wrf_python_tutorial git repository linked to in Step 5 includes a directory containing several WRF-ARW data files which will be used for examples during the tutorial.

You also have the option of using your own data files for the tutorial by modifying the first Jupyter Notebook cell to point to your data set. However, there is no guarantee that every cell in your workbook will work without some modifications (e.g. cross section lines will be drawn outside of your domain).

If you have recently cloned the wrf_python_tutorial git repository, then you should have a “wrf_tutorial_data” directory at the root level of the “wrf_python_tutorial” directory.
If this directory does not exist, try running the following commands from within the “wrf_python_tutorial” directory to update your local copy of the git repository:
git checkout -- .
git pull
git submodule init
git submodule update
cd wrf_tutorial_data
git checkout -- .
Verify that you have three WRF output files in the “wrf_tutorial_data” directory:
$ ls wrf_tutorial_data
wrfout_d01_2005-08-28_00_00_00
wrfout_d01_2005-08-28_12_00_00
wrfout_d01_2005-08-29_00_00_00

Getting Help¶

If you experience problems during this installation, please send a question to the Google Group support mailing list.

We look forward to seeing you at the tutorial!

Past Tutorials¶

WRF-Python and VAPOR Workshop 2018 (Boise State University)¶

The Department of Geosciences at Boise State University is partnering with staff from the National Center for Atmospheric Research (NCAR) to host a free, 2-day workshop in the Environmental Research Building (ERB) lab 2104 at Boise State University on September 26-27, 2018. The tutorial will be centered on the WRF-Python and VAPOR tools for analyzing and visualizing data from the Weather Research and Forecasting (WRF) regional weather and climate model.

Users must be registered to attend this tutorial (see Registration).

Location¶

September 26-27, 2018 9:00 AM - 4:00 PM

Boise State University, Environmental Research Building (ERB) lab #2104.

WRF-Python Overview¶

WRF-Python is a collection of diagnostic and interpolation routines for use with output from the Weather Research and Forecasting (WRF-ARW) Model. The package provides over 30 diagnostic calculations, several interpolation routines, and utilities to help with plotting via cartopy, basemap, or PyNGL. The functionality is similar to what is provided by the NCL WRF package.

Note

WRF-Python is NOT a tool for running the WRF-ARW model using Python.

This tutorial provides an introduction to wrf-python. The tutorial is beginner friendly for new users of wrf-python, but this is NOT an introduction to the Python programming language (see Prerequisites). Due to limited seating, if you do not have any previous experience with Python, please do not register for this tutorial.

Note

For online training that provides an introduction to the Python programming language itself, please see the Unidata Python Training Page.

Computers will be provided, but feel free to use your own laptop if you prefer. We will be covering how to install wrf-python via conda as part of the tutorial.

Students are encouraged to bring their own data sets, but data will be provided if this is not an option. Students will be provided a jupyter notebook workbook which can be modified to accommodate their data.

Topics include:

How to install wrf-python via conda
A brief introduction to jupyter notebook
Overview of WRF data files
WRF-Python basics
Plotting with cartopy
Overview of OpenMP features and other performance tips
Open lab for students

Registration¶

Please register prior to September 19, 2018. The registration form is here:

Registration Form

Registration consists of a brief survey, which will help give the instructor a brief overview of your background and will help tailor the tutorial to your expectations.

Prerequisites¶

This tutorial assumes that you have basic knowledge of how to type commands in to a terminal window using your preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.

This tutorial assumes that you have prior experience programming in Python. Below is a list of some Python concepts that you will see in the examples, but don’t worry if you aren’t familiar with everything.

Opening a Python interpreter and entering commands.
Importing packages via the import statement.
Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
Calling functions (e.g. result = some_function(x, y))
Familiarity with numpy would be helpful, as only a very brief introduction is provided.
Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

Instructions for Computer Lab Installation¶

Step 1: Download Miniconda¶

For this tutorial, you will need to download and install Miniconda. We are going to use Python 3.6+.

Please use the appropriate link below to download Miniconda for your operating system.

Note

64-bit OS recommended

https://www.learnpython.org/

For more information, see: https://conda.io/miniconda.html

Step 2: Install Miniconda¶

Windows:

Browse to the directory where you downloaded Miniconda3-latest-Windows-x86_64.exe.

Double click on Miniconda3-latest-Windows-x86_64.exe.

Follow the instructions.

For Windows 10, use the Anaconda command prompt found under the Anaconda2 menu (Start Menu -> Anaconda2 -> Anaconda Prompt). Otherwise, open a regular command prompt.

Mac and Linux:

For Mac and Linux, the installer is a bash script.
Using a terminal, you need to execute the bash shell script that you downloaded by doing:
bash /path/to/Miniconda3-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda3-latest-Linux-x86_64.sh [Linux]
Follow the instructions.
At the end of the installation, it will ask if you want to add the miniconda3 path to your bash environment. If you are unsure what to do, you should say “yes”. If you say “no”, we’re going to assume you know what you are doing.

If you said “yes”, then once you restart your shell, the miniconda3 Python will be found instead of the system Python when you type the “python” command. If you want to undo this later, then you can edit either ~/.bash_profile or ~/.bashrc (depending on OS used) and comment out the line that looks similar to:
# added by Miniconda3 x.x.x installer
export PATH="/path/to/miniconda3/bin:$PATH"
Restart your command terminal.
[Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then you need to activate the bash shell by typing the following in to your command terminal:
bash
Verify that your system is using the correct Python interpreter by typing the following in to your command terminal:
which python
You should see the path to your miniconda installation. If not, see the note below.

Note

If you have already installed another Python distribution, like Enthought Canopy, you will need to comment out any PATH entries for that distribution in your .bashrc or .bash_profile. Otherwise, your shell environment may pick to wrong Python installation.

If bash is not your default shell type, and the PATH variable has been set in .bash_profile by the miniconda installer, try executing “bash -l” instead of the “bash” command in step 5.

Step 3: Set Up the Conda Environment¶

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when working with conda.

The name of our conda environment for this tutorial is: tutorial_backup.

Follow the instructions below to create the tutorial_backup environment.

Open a command terminal if you haven’t done so.
[Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not your current shell, type:
bash
Add the conda-forge channel to your conda package manager.

Type or copy the command below in to your command terminal. You should run this command even if you have already done it in the past. This will ensure that conda-forge is set as the highest priority channel.
conda config --add channels conda-forge
Note

Conda-forge is a community driven collection of packages that are continually tested to ensure compatibility. We highly recommend using conda-forge when working with conda. See https://conda-forge.github.io/ for more details on this excellent project.
Create the backup conda environment for the tutorial.

Students will create a conda environment during the tutorial, but if they run in to problems, we’re going to create a backup environment.

Type or copy this command in to your command terminal:
conda create -n tutorial_backup python=3.6 matplotlib cartopy netcdf4 jupyter git ffmpeg wrf-python
Type “y” when prompted. It will take several minutes to install everything.

This command creates an isolated Python environment named tutorial_backup, and installs the python interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.
Note

When the installation completes, your command terminal might post a message similar to:
If this is your first install of dbus, automatically load on login with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda3/envs/tutorial_test/org.freedesktop.dbus-session.plist ~/Library/LaunchAgents/
launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-session.plist
This is indicating that the dbus package can be set up to automatically load on login. You can either ignore this message or type in the commands as indicated on your command terminal. The tutorial should work fine in either case.
Activate the conda environment.

To activate the tutorial_backup Python environment, type the following in to the command terminal:

For Linux and Mac (using bash):
source activate tutorial_backup
For Windows:
activate tutorial_backup
You should see (tutorial_backup) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:

For Linux and Mac:
source deactivate
For Windows:
deactivate tutorial_backup

Step 4: Download the Student Workbook¶

The student workbook for the tutorial is available on GitHub. The tutorial_backup conda environment includes the git application needed to download the repository.

These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another directory, we’re going to assume you know how to do this yourself.

To download the student workbook, follow these instructions:

Activate the tutorial_backup conda environment following the instructions in the previous step (source activate tutorial_backup or activate tutorial_backup).
Change your working directory to the home directory by typing the following command in to the command terminal:

For Linux and Mac:
cd ~
For Windows:
cd %HOMEPATH%
Download the git repository for the tutorial by typing the following in to the command terminal:
git clone https://github.com/NCAR/wrf_python_tutorial.git
There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes, type the following in to the command terminal:

For Linux and Mac:
source activate tutorial_backup

cd ~/wrf_python_tutorial/boise_workshop_2018

git pull
For Windows:
activate tutorial_2018

cd %HOMEPATH%\wrf_python_tutorial\boise_workshop_2018

git pull
Note

If you try the “git pull” command and it returns an error indicating that you have made changes to the workbook, this is probably because you ran the workbook and it contains the cell output. To fix this, first do a checkout of the workbook, then do the pull.
git checkout -- .
git pull

Step 5: Verify Your Environment¶

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see some warnings for matplotlib or xarray, but you can safely ignore these).

Activate the tutorial_backup conda environment if it isn’t already active (see instructions above).
Open a python terminal by typing the following in to the command terminal:
python
Now type the following in to the Python interpreter:
>>> import netCDF4
>>> import matplotlib
>>> import xarray
>>> import wrf
You can exit the Python interpreter using CTRL + D

Step 6: Install WRF Output Files¶

A link will be provided in an email prior to the tutorial for the WRF-ARW data files used for the examples.

The link in the email should take you to a location on an Amazon cloud drive.

If you hover your mouse over the wrf_tutorial_data.zip file, you’ll see an empty check box appear next to the file name. Click this check box.

At the bottom of the screen, you’ll see a Download button next to a cloud icon. Click this button to start the download.

The download was most likely placed in to your ~/Downloads folder [%HOMEPATH%\Downloads for Windows]. Using your preferred method of choice for unzipping files, unzip this file in to your home directory. Your data should now be in ~/wrf_tutorial_data [%HOMEPATH%\wrf_tutorial_data for Windows].

Verify that you have three WRF output files in that directory.

WRF Users’ Workshop 2018¶

Welcome WRF-Python Tutorial attendees!

If you wish to actively participate in the tutorial, please bring your own laptop. Due to limited time constraints, the instructions below should be completed prior to arriving at the tutorial.

I will be executing the same cells as the student workbook, so if you prefer to sit and watch, that’s OK too. Following the tutorial, I will upload the instructor slides in to the same GitHub location as the student workbook if you want to try it out later.

Prerequisites¶

This tutorial assumes that you have basic knowledge of how to type commands in to a command terminal using your preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.

Regarding Python, to understand the examples in this tutorial, you should have some experience with Python basics. Below is a list of some Python concepts that you will see in the examples, but don’t worry if you aren’t familiar with everything.

Opening a Python interpreter and entering commands.
Importing packages via the import statement.
Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
Calling functions (e.g. result = some_function(x, y))
Familiarity with numpy would be helpful, as only a very brief introduction is provided.
Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

If you are completely new to Python, that shouldn’t be a problem, since most of the examples consist of basic container types and function calls. It would be helpful to look at some introductory material before arriving at the tutorial. If you’ve programmed before, picking up Python isn’t too difficult.

Here are some links:

https://developers.google.com/edu/python/

Step 1: Open a Command Terminal¶

To begin, you will first need to know how to open a command line terminal for your operating system.

For Windows:

WINDOWS + r
type cmd in the run window

For Mac:

Finder -> Applications -> Utilities -> Terminal

For Linux:

Try one of the following:

CTRL + ALT + T
CTRL + ALT + F2

Step 2: Download Miniconda¶

For this tutorial, you will need to download and install Miniconda. We are going to use Python 3.6, but it will also work with Python 2.7.

Please use the appropriate link below to download Miniconda for your operating system.

Note

64-bit OS recommended

https://www.learnpython.org/

For more information, see: https://conda.io/miniconda.html

Note

What is Miniconda?

If you have used the Anaconda distribution for Python before, then you will be familiar with Miniconda. The Anaconda Python distribution includes numerous scientific packages out of the box, which can be difficult for users to build and install. More importantly, Anaconda includes the conda package manager.

The conda package manager is a utility (similar to yum or apt-get) that installs packages from a repository of pre-compiled Python packages. These repositories are called channels. Conda makes it easy for Python users to install and uninstall packages, and also can be used to create isolated Python environments (more on that later).

Miniconda is a bare bones implementation of Anaconda and only includes the conda package manager. Since we are going to use the conda-forge channel to install our scientific packages, Miniconda avoids any complications between packages provided by Anaconda and conda-forge.

Step 3: Install Miniconda¶

Windows:

Browse to the directory where you downloaded Miniconda3-latest-Windows-x86_64.exe.

Double click on Miniconda3-latest-Windows-x86_64.exe.

Follow the instructions.

Restart your command terminal.

Mac and Linux:

For Mac and Linux, the installer is a bash script.
Using a terminal, you need to execute the bash shell script that you downloaded by doing:
bash /path/to/Miniconda3-latest-MacOSX-x86_64.sh [Mac]

bash /path/to/Miniconda3-latest-Linux-x86_64.sh [Linux]
Follow the instructions.
At the end of the installation, it will ask if you want to add the miniconda3 path to your bash environment. If you are unsure what to do, you should say “yes”. If you say “no”, we’re going to assume you know what you are doing.

If you said “yes”, then once you restart your shell, the miniconda3 Python will be found instead of the system Python when you type the “python” command. If you want to undo this later, then you can edit either ~/.bash_profile or ~/.bashrc (depending on OS used) and comment out the line that looks similar to:
# added by Miniconda3 x.x.x installer
export PATH="/path/to/miniconda3/bin:$PATH"
Restart your command terminal.
[Linux and Mac Users Only] Miniconda only works with bash. If bash is not your default shell, then you need to activate the bash shell by typing the following in to your command terminal:
bash
Verify that your system is using the correct Python interpreter by typing the following in to your command terminal:
which python
You should see the path to your miniconda installation. If not, see the note below.

Note

If you have already installed another Python distribution, like Enthought Canopy, you will need to comment out any PATH entries for that distribution in your .bashrc or .bash_profile. Otherwise, your shell environment may pick to wrong Python installation.

If bash is not your default shell type, and the PATH variable has been set in .bash_profile by the miniconda installer, try executing “bash -l” instead of the “bash” command in step 5.

Step 4: Set Up the Conda Environment¶

If you are new to the conda package manager, one of the nice features of conda is that you can create isolated Python environments that prevent package incompatibilities. This is similar to the virtualenv package that some Python users may be familiar with. However, conda is not compatible with virtualenv, so only use conda environments when working with conda.

The name of our conda environment for this tutorial is: tutorial_2018.

Follow the instructions below to create the tutorial_2018 environment.

Open a command terminal if you haven’t done so.
[Linux and Mac Users Only] The conda package manager only works with bash, so if bash is not your current shell, type:
bash
Add the conda-forge channel to your conda package manager.

Type or copy the command below in to your command terminal. You should run this command even if you have already done it in the past. This will ensure that conda-forge is set as the highest priority channel.
conda config --add channels conda-forge
Note

Conda-forge is a community driven collection of packages that are continually tested to ensure compatibility. We highly recommend using conda-forge when working with conda. See https://conda-forge.github.io/ for more details on this excellent project.
Create the conda environment for the tutorial.

Type or copy this command in to your command terminal:
conda create -n tutorial_2018 python=3.6 matplotlib cartopy netcdf4 jupyter git ffmpeg wrf-python
Type “y” when prompted. It will take several minutes to install everything.

This command creates an isolated Python environment named tutorial_2018, and installs the python interpreter, matplotlib, cartopy, netcdf4, jupyter, git, ffmpeg, and wrf-python packages.
Note

When the installation completes, your command terminal might post a message similar to:
If this is your first install of dbus, automatically load on login with:

mkdir -p ~/Library/LaunchAgents
cp /path/to/miniconda3/envs/tutorial_test/org.freedesktop.dbus-session.plist ~/Library/LaunchAgents/
launchctl load -w ~/Library/LaunchAgents/org.freedesktop.dbus-session.plist
This is indicating that the dbus package can be set up to automatically load on login. You can either ignore this message or type in the commands as indicated on your command terminal. The tutorial should work fine in either case.
Activate the conda environment.

To activate the tutorial_2018 Python environment, type the following in to the command terminal:

For Linux and Mac (using bash):
source activate tutorial_2018
For Windows:
activate tutorial_2018
You should see (tutorial_2018) on your command prompt.

To deactivate your conda environment, type the following in to the command terminal:

For Linux and Mac:
source deactivate
For Windows:
deactivate tutorial_2018

Step 5: Download the Student Workbook¶

The student workbook for the tutorial is available on GitHub. The tutorial_2018 conda environment includes the git application needed to download the repository.

These instructions download the tutorial in to your home directory. If you want to place the tutorial in to another directory, we’re going to assume you know how to do this yourself.

To download the student workbook, follow these instructions:

Activate the tutorial_2018 conda environment following the instructions in the previous step (source activate tutorial_2018 or activate tutorial_2018).
Change your working directory to the home directory by typing the following command in to the command terminal:

For Linux and Mac:
cd ~
For Windows:
cd %HOMEPATH%
Download the git repository for the tutorial by typing the following in to the command terminal:
git clone https://github.com/NCAR/wrf_python_tutorial.git
There may be additional changes to the tutorial after you have downloaded it. To pull down the latest changes, type the following in to the command terminal:

For Linux and Mac:
source activate tutorial_2018

cd ~/wrf_python_tutorial/wrf_workshop_2018

git pull
For Windows:
activate tutorial_2018

cd %HOMEPATH%\wrf_python_tutorial\wrf_workshop_2018

git pull
Note

If you try the “git pull” command and it returns an error indicating that you have made changes to the workbook, this is probably because you ran the workbook and it contains the cell output. To fix this, first do a checkout of the workbook, then do the pull.
git checkout -- .
git pull

Step 6: Verify Your Environment¶

Verifying that your environment is correct involves importing a few packages and checking for errors (you may see some warnings for matplotlib or xarray, but you can safely ignore these).

Activate the tutorial_2018 conda environment if it isn’t already active (see instructions above).
Open a python terminal by typing the following in to the command terminal:
python
Now type the following in to the Python interpreter:
>>> import netCDF4
>>> import matplotlib
>>> import xarray
>>> import wrf
You can exit the Python interpreter using CTRL + D

Step 7: Obtain WRF Output Files¶

A link will be provided in an email prior to the tutorial for the WRF-ARW data files used for the examples. If you did not receive this email, the link will also be provided at the tutorial itself.

You also have the option of using your own data files for the tutorial by modifying the first Jupyter Notebook cell to point to your data set. However, there is no guarantee that every cell in your workbook will work without some modifications (e.g. cross section lines will be drawn outside of your domain).

The link in the email should take you to a location on an Amazon cloud drive.

If you hover your mouse over the wrf_tutorial_data.zip file, you’ll see an empty check box appear next to the file name. Click this check box.

At the bottom of the screen, you’ll see a Download button next to a cloud icon. Click this button to start the download.

The download was most likely placed in to your ~/Downloads folder [%HOMEPATH%\Downloads for Windows]. Using your preferred method of choice for unzipping files, unzip this file in to your home directory. Your data should now be in ~/wrf_tutorial_data [%HOMEPATH%\wrf_tutorial_data for Windows].

Verify that you have three WRF output files in that directory.

Getting Help¶

If you experience problems during this installation, please send a question to the Google Group support mailing list.

We look forward to seeing you at the tutorial!

WRF-Python Tutorial 2018¶

NCAR will be providing a four hour tutorial for wrf-python on Wednesday, March 7, 2018 from 8 AM to 12 PM. The tutorial is free, but seating is limited to only 16 students, so registration is required.

The tutorial will take place at NCAR’s corporate training center in Boulder, Colorado.

Corporate Technical Training Center 3085 Center Green Drive, Building CG-2, Room #3024 Boulder, Colorado

Overview¶

WRF-Python is a collection of diagnostic and interpolation routines for use with output from the Weather Research and Forecasting (WRF-ARW) Model. The package provides over 30 diagnostic calculations, several interpolation routines, and utilities to help with plotting via cartopy, basemap, or PyNGL. The functionality is similar to what is provided by the NCL WRF package.

Note

WRF-Python is NOT a tool for running the WRF-ARW model using Python.

This tutorial provides an introduction to wrf-python. The tutorial is beginner friendly for new users of wrf-python, but this is NOT an introduction to the Python programming language (see Prerequisites). Due to limited seating, if you do not have any previous experience with Python, please do not register for this tutorial.

Note

For online training that provides an introduction to the Python programming language itself, please see the Unidata Python Training Page.

Computers will be provided, but feel free to use your own laptop if you prefer. We will be covering how to install wrf-python via conda as part of the tutorial.

Students are encouraged to bring their own data sets, but data will be provided if this is not an option. Students will be provided a jupyter notebook workbook which can be modified to accommodate their data.

Topics include:

How to install wrf-python via conda
A brief introduction to jupyter notebook
Overview of WRF data files
WRF-Python basics
Plotting with cartopy
Overview of OpenMP features and other performance tips
Open lab for students

Registration¶

Please register prior to February 21, 2018. The registration form is here:

Registration Form

Registration consists of a brief survey, which will help give the instructor a brief overview of your background and will help tailor the tutorial to your expectations.

Prerequisites¶

This tutorial assumes that you have basic knowledge of how to type commands in to a terminal window using your preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.

This tutorial assumes that you have prior experience programming in Python. Below is a list of some Python concepts that you will see in the examples, but don’t worry if you aren’t familiar with everything.

Opening a Python interpreter and entering commands.
Importing packages via the import statement.
Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
Calling functions (e.g. result = some_function(x, y))
Familiarity with numpy would be helpful, as only a very brief introduction is provided.
Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

WRF Users’ Workshop 2017¶

Welcome wrf-python tutorial attendees!

The instructions below should be completed prior to arriving at the tutorial. There will not be enough time to do this during the tutorial.

Prerequisites¶

This tutorial assumes that you have basic knowledge of how to type commands in to a command terminal using your preferred operating system. You should know some basic directory commands like cd, mkdir, cp, mv.

Regarding Python, to understand the examples in this tutorial, you should have some experience with Python basics. Below is a list of some Python concepts that you will see in the examples, but don’t worry if you aren’t familiar with everything.

Opening a Python interpreter and entering commands.
Importing packages via the import statement.
Familiarity with some of the basic Python types: str, list, tuple, dict, bool, float, int, None.
Creating a list, tuple, or dict with “[ ]”, “( )”, “{ }” syntax (e.g. my_list = [1,2,3,4,5]).
Accessing dict/list/tuple items with the “x[ ]” syntax (e.g. my_list_item = my_list[0]).
Slicing str/list/tuple with the “:” syntax (e.g. my_slice = my_list[1:3]).
Using object methods and attributes with the “x.y” syntax (e.g. my_list.append(6)).
Calling functions (e.g. result = some_function(x, y))
Familiarity with numpy would be helpful, as only a very brief introduction is provided.
Familiarity with matplotlib would be helpful, as only a very brief introduction is provided.

If you are completely new to Python, that shouldn’t be a problem, since most of the examples consist of basic container types and function calls. It would be helpful to look at some introductory material before arriving at the tutorial. If you’ve programmed before, picking up Python isn’t too difficult.

Here are some links:

https://developers.google.com/edu/python/

Step 1: Open a Command Terminal¶

To begin, you will first need to know how to open a command line terminal for your operating system.

For Windows:

WINDOWS + r
type cmd in the run window

For Mac:

Finder -> Applications -> Utilities -> Terminal

For Linux:

Try one of the following:

CTRL + ALT + T
CTRL + ALT + F2

Step 2: Download Miniconda¶

For this tutorial, you will need to download and install Miniconda. We are going to use Python 2.7, but it should also work with Python 3.5+. However, due to limitations with open source compilers on conda-forge, only Python 2.7 is available for Windows.

Please use the appropriate link below to download Miniconda for your operating system.

Note

64-bit OS recommended