Running locally with Python

Scripts

The scripts/ folder contains some simple scripts to help with development. To run them, ensure they are made executable in your filesystem (they may not be by default depending on your OS).

You can do that in whichever File > Permissions > Make Executable menu your desktop provides.

For *nix environments or the WSL, you can type chmod +x <filename> to add executable permissions.

Run all scripts from the root of the project, or they won't work.

Running the dGC Server locally with Python

Note

Some of this setup is obvious to experienced Python developers, but it's documented here so we all know the same obvious . This helps us reduce development difficulty and speeds up onboarding of new team members.

Managing Python versions, and dependencies such as libraries

Managing Python versions

Currently, we use Python 3.8.3 for these algorithms.

There are different tools available to help you manage multiple different Python versions on the same machine. We use pyenv here, however, there are other ways to solve this problem. If you already have a preferred method, you should be able to use that.

Managing library / dependencies versions

If you pip install every dependency in requirements.txt globally on your machine, you can encounter problems if you develop other Python applications on the same machine. For example, different projects may need different versions of the same library.

Our solutions are:

If using Mac / Linux, use pyenv-virtualenv which is an extension to pyenv which helps you to manage separate 'environments' for each Python project you work on.
If using Windows, use virtualenv, which is a popular too to create isolated Python environments for Python libraries.

Please see this StackOverflow post to find out more about the differences.

Reason for not using pyenv on Windows

pyenv does not directly support Windows. There is a Windows port in development, however, it is simpler to just use virtualenv (and also leads to fewer headaches setting up the development environment!).

Mac/Linux - installing `pyenv`

pyenv installer

After installing and setting up pyenv, the correct Python version will be automatically selected when you navigate to the directory containing this repository, because of the .python-version file.

Example setup commands for this repository

git clone this repository into a suitable location on your development machine

git clone https://github.com/rcpch/digital-growth-charts-server.git

cd into the directory

cd digital-growth-charts-server

Install the correct Python version

pyenv install 3.8.0

Create a virtualenv for this project 'growth-charts', abbreviated to 'gc-3.8' using Python 3.8.0

pyenv virtualenv 3.8.0 dgc-server

Auto-selection of Python and virtualenv

Using 'dgc-server' as the name will enable it to be automatically selected when navigating to this repo (but you can call your own virtualenv whatever you like). This all works using the .python-version file in the project root. This can contain either a Python version name which pyenv recognises, or it can contain a virtualenv name, which pyenv will select for you (and this automatically selects the Python version too).

A helpful article about this is here: https://realpython.com/intro-to-pyenv/#activating-your-versions.

Check virtualenv creation worked

pyenv virtualenvs should return something like:

dgc-server (created from /home/my-user/.pyenv/versions/3.8.0)

Activate the virtualenv manually if it's not already selected

pyenv activate dgc-server

Install the dependencies inside this virtualenv

pip install -r requirements.txt

Refer to the pyenv command reference if you need further information on pyenv

Extra development packages that may be required on some setups

On some platforms, you may need the additional development header packages. On Ubuntu/Linux Mint this was required when using pyenv and thus compiling Python from source. This should not be necessary if you're running a binary Python, it only affects setups which are compiling a specific Python version from source, on demand, such as pyenv.

sudo apt-get install liblzma-dev libbz2-dev zlib1g-dev

and then recompile the Python that pyenv built earlier

pyenv install 3.8.3

If installing on macOS Big Sur, pyenv install of python 3.8.0 and requirements.txt may fail

To install 3.8.3 via pyenv, set the following 2 environment variables (requires homebrew installed versions of bzip2, openssl and zlib):

export CFLAGS="-I$(brew --prefix openssl)/include -I$(brew --prefix bzip2)/include -I$(brew --prefix readline)/include -I$(xcrun --show-sdk-path)/usr/include"
export LDFLAGS="-L$(brew --prefix openssl)/lib -L$(brew --prefix readline)/lib -L$(brew --prefix zlib)/lib -L$(brew --prefix bzip2)/lib"

Now, run the pyenv install with a patch for Big Sur:

pyenv install --patch 3.8.0 < <(curl -sSL https://github.com/python/cpython/commit/8ea6353.patch\?full_index\=1)

Now, once ready to install requirements.txt with pip, set one more environment variable:

export SYSTEM_VERSION_COMPAT=1

Skip the following Windows section, to Start the API server natively with default settings once complete.

Windows - installing `virtualenv`

git clone this repository into a suitable location on your development machine

git clone https://github.com/rcpch/digital-growth-charts-server.git

cd into the directory

cd digital-growth-charts-server

First, ensure you update pip, then use it to install virtualenv

python.exe -m pip install --upgrade pip

pip install virtualenv

Create a virtual environment called env (or any name you want - but make sure to reference the correct name going forwards)

py -m venv env

Navigate to the /Scripts folder

cd env/Scripts

Run activate.bat

activate.bat

You should then see the name of your virtual environment prepend your prompts e.g.

(env) C:\Users\...\digital-growth-charts-server\env\Scripts>

Now, go back to the root directory

(env) C:\Users\...\digital-growth-charts-server\env\Scripts> cd ..
(env) C:\Users\...\digital-growth-charts-server\env\> cd ..

And install the dependences e.g.

(env) C:\Users\...\digital-growth-charts-server> pip install -r requirements/common-requirements.txt

Start the API server natively with default settings

From the application's root directory, type

Mac/Linux

s/uvicorn-start

Windows

uvicorn main:app --reload

You should see messages from the uvicorn development server like:

INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
INFO:     Started reloader process [61645] using watchgod
INFO:     Started server process [61647]
INFO:     Waiting for application startup.
INFO:     Application startup complete.

There may be other messages at the end of the output for other processes which run on server start-up.

If you need to vary any of the parameters passed, you can either:

Modify the start-up script
Manually pass the commands to the shell, using the commands in the start-up script as a guide