Add first rough draft

slides.qmd

@@ -11,4 +11,157 @@ author:
   - name: Lukas Kluft
 ---
 
-# Of pythons and other snakes
+# Python
+
+* Invented by Guido van Rossum in XXXX
+* Usually considered to have a simple syntax
+* Interpreted scripting lanuage
+* De-facto standard in many data science fields
+
+## Interpreter
+
+* Python is not compiled^[this is true but also not true] but interpreted on runtime
+* The interpreter can be called directly
+  ```{raw}
+  python3 my_scripy.py
+  ```
+* Scripts can define an interpreter in the shebang
+  ```
+  #!/usr/bin/env python3
+  ```
+
+# Standard library
+
+* Python organizes code in individual packages
+* [Standard library](https://docs.python.org/3/library/index.html) provide extensive built-in facilities (`os`, `sys`, `datetime`, `itertools`, ...)
+
+# External packages {auto-animate=true}
+
+* The python package index ([PyPI](https://pypi.org)) is a centralized repository for third-party packages
+* Everyone can register packages under a not yet taken name (first come first serve)
+* Packages can be installed using the built-in package manager pip:
+  ```{raw}
+  python3 -m pip install numpy
+  ```
+
+# Package manager
+
+* A package manager is a software that helps you to manage different version of packages and their dependencies
+* General purpose package manager (`brew`, `apt-get`, `conda`, ...)
+* Python package manager (`pip`, `conda`, ...)
+
+# _Virtual_ Environments
+
+* Manage different sets of python packages
+* Allows you to have a stable combination of package versions
+* Pure python tools (`virtualenv`, ...)
+* Tools to handle general software environments (`conda`, ...)
+
+# (Ana)conda
+
+* `conda` is a general purpose package manager provided by Anacond Inc.
+* `conda` allows you to install
+  1. handle virtual environments
+  2. with different python interpreters
+  3. and install python packages
+
+# Fastest snakes on earth
+
+:::: {.columns}
+
+::: {.column width="30%"}
+Python
+
+* Foo
+* Bar
+:::
+
+::: {.column width="30%"}
+Anaconda
+
+* Very big
+* Slow
+:::
+
+::: {.column width="30%"}
+Mamba
+
+* Venomous
+* Fastest snake
+:::
+
+::::
+
+# (micro)mamba
+
+* `mamba` is a drop-in replacement for `conda` written in C++
+* Conflict resolution is **much** faster
+* Can be installed without any pre-occupied environment (`micromamba`)
+
+# Setting up an environment
+
+```{raw}
+mamba create -n my_env python=3.12
+mamba activate my_env
+mamba install matplotlib numpy scipy
+```
+
+# Recreating an environment {auto-animate=true visibility="uncounted"}
+
+Specify environment name, python version and dependencies:
+```{.yaml filename=environment.yaml}
+name: my_env
+channels:
+  - conda-forge
+dependencies:
+  - matplotlib
+  - numpy
+  - scipy
+```
+
+Create a new environment based on your configuration:
+```{raw}
+mamba env create -f environment.yaml
+mamba activate my_env
+```
+
+# Recreating an environment {auto-animate=true}
+
+Specify environment name, python version and dependencies:
+```{.yaml filename=environment.yaml}
+name: my_env
+channels:
+  - conda-forge
+dependencies:
+  - matplotlib
+  - numpy
+  - scipy
+  - pip:
+    - gribscan
+```
+
+Create a new environment based on your configuration:
+```{raw}
+mamba env create -f environment.yaml
+mamba activate my_env
+```
+
+# Excursion: Ipython kernels
+
+* Make an environment usable in Jupyterhub
+* Install `ipykernel` package to manage different kernels:
+  ```{raw}
+  python3 -m pip install ipykernel
+  ```
+  _(alternatively specify in your `environment.yaml`)_
+* Install the currently active (!) environment as kernel:
+  ```{raw}
+  python3 -m ipykernel install --name my_env
+  ```
+* The current environment should be available as kernel in Jupyter
+
+# Conclusions
+
+* Use virtual environments to organize your work
+* Keep track of installed packages in an `environment.yaml` file
+* If in doubt, use `micromamba` to do so