Update installing_Python.md

Author: marcoalopez

notebooks/installing_Python.md

@@ -22,22 +22,22 @@ For this course, we will use Miniconda which allows us to manage Python packages
 4. First, make sure you have the latest version of **conda** installed. To do this, type the following command:
 
    ```
-   > conda update conda
+   conda update conda
    ```
    and update if necessary (👉 tip: do this often). 
 
 5. If you have been paying attention, you will see that the current line in the console says "base" in parentheses. This means that you are in a Python environment called "base". A wise strategy is to not install anything in this environment other than the standard Python libraries that have already been added during the Miniconda installation process. If you are curious, you can see which libraries these are by entering the command ``conda list``. Our strategy here will be to create a new Python environment  called "course" and install there all the scientific libraries that we will use during the course. To do this, we will enter the following command:
 
    ```
-   > conda create --name course python
+   conda create --name course python
    ```
 
 6. Once finished, use the command ``conda env list`` and you will see two environments listed, "base" and "course", and an asterisk indicating that "base" is the currently active environment. To install the Python libraries needed in the "course" environment, we need to activate it first, so we type ``activate course``. Now you will see the word "course" in parentheses, indicating that "course" is now the active environment.
 
 7. Next, we proceed to install the libraries needed for the course. Instead of doing it one by one, we will do it in a single line as follows (you can paste and copy the command directly into your console)
 
    ```
-   >conda install numpy scipy pandas matplotlib jupyterlab
+   conda install numpy scipy pandas matplotlib jupyterlab
    ```
 
    and say yes to all. Conda will install the chosen libraries/packages and all the necessary dependencies. Once finished, you're done.
@@ -71,33 +71,31 @@ We will learn how to use Jupyter Lab properly during the course.
 The following are the basic commands for installing, removing and and keep your Python libraries up to date.
 
 ```markdown
-list all Python packages and the versions installed in a specific environment:
-> conda list
+# list all Python packages and the versions installed in a specific environment:
+conda list
 
-install a new package in a specific environment:
-> conda install <name(s) of the package(s)>
+# install a new package in a specific environment:
+conda install <name of the package-s>
 
-uninstall a package
-> conda remove <name of the package>
+# uninstall a package
+conda remove <name of the package>
 
-update a specific package
-> conda update <name of the package>
+# update a specific package
+conda update <name of the package>
 
-update all packages in the environment ensuring compatibility
-> conda update --all
+# update all packages in the environment ensuring compatibility
+conda update --all
 
-clean all outdated packages (by default conda does not delete
-previous versions of packages/libraries when upgrading).
-> conda clean --all
+# clean all outdated packages (by default conda does not delete
+# previous versions of packages/libraries when upgrading).
+conda clean --all
 ```
 
 
 
 
 ## More on conda/Python environments
 
-TODO
-
 Environments enables you to have multiple versions of Python or scientific packages installed on your computer at the same time. This is useful, for example, to test a script on different versions of Python and any other Python library, or to keep separate different scientific libraries that may work with different Python versions, etc.
 
 ```
@@ -116,25 +114,56 @@ to create a new environment using conda the general procedure is as follows (in
 some examples below
 
 ```markdown
-create an environment named "main" with the last version of Python supported by conda
->conda create --name main python
+# create an environment named "main" with the last version of Python supported by conda
+conda create --name main python
 
-create an environment with a specific version of Python
->conda create --name new_env python=3.8.1
+# create an environment with a specific version of Python
+conda create --name new_env python=3.8.1
 
-create an environment named "SCIENV" with the libraries numpy, scipy, matplolib and jupyterlab (conda will take care of including all the necessary dependencies)
->conda create --name SCIENV numpy scipy matplotlib jupyterlab
+# create an environment named "SCIENV" with the libraries numpy, scipy, matplolib and jupyterlab (conda will take care of including all the necessary dependencies)
+conda create --name SCIENV numpy scipy matplotlib jupyterlab
 
-create an environment named "image" with the library scikit-image and all the neccesary dependencies
->conda create --name image scikit-image
+# create an environment named "image" with the library scikit-image and all the neccesary dependencies
+conda create --name image scikit-image
 
-remove an existing environment (make sure you are not in the environment you want to remove)
->conda env remove --name <ENV NAME>
+# remove an existing environment (make sure you are not in the environment you want to remove)
+conda env remove --name <ENV NAME>
 ```
 
+More info here: https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html
+
+### Adding Python packaged from other channels or via pip in conda
 
+At some point, you are likely to find yourself in the situation where you want to install a Python package that is not in the official (default) conda channel, and you need to install it using pip, the standard Python installation package, or the community-driven _conda-forge_ channel. In this case, it is strongly recommended to create a new Python environment and, in particular, never use pip to install packages into the Python environments you have for general use, as this may break compatibility between different versions of the environment.
+
+#### Case 1: Python package installed using pip
+
+Create a new Python environment and install the package and any necessary dependencies. Follow the developer's installation instructions.
+
+#### Case 2: the Python package is in the conda-forge channel
+
+To install packages from the _conda-forge_ community channel, you must first activate it (see instructions below). Once this is done, create your new Python environment and install all the necessary libraries using conda.
+
+To add the _conda-forge_ channel use:
+
+```markdown
+conda config --append channels conda-forge
+```
+
+> [!WARNING]
+> In most tutorials you can find on the internet you will find the following line to enable the conda-forge channel: ``conda config --add channels conda-forge``. **We advise you not to do this**. This will make conda-forge the priority channel in your Python installation. Your anaconda's default channel will have a lower priority, and some of your packages will start upgrading to conda-forge, affecting other environments as well. Using ``--append`` instead of ``--add`` will keep your default conda channel high in priority, and package versions will be searched on your default before going to conda-forge. If you have already made a mess of it, you can use the following lines:
+>
+> ```markdown
+> # show channels
+> conda config --show channels
+> 
+> # describe channel priority
+> conda config --describe channel_priority
+> 
+> # remove conda-forge
+> conda config --remove channels conda-forge
+> ```
 
-More info here: https://docs.conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html
 
 ### Using environment.ylm (YALM) files
 
@@ -146,15 +175,16 @@ These files are a useful way to share Python environments and ensure, for exampl
 
 To create a conda environment using an ``environment.ylm`` file from the command lines use (if you are in the folder containing the ``environment.ylm`` otherwise you will have to specify the path)
 
-``>conda env create -f environment.yml``
+``conda env create -f environment.yml``
 
 This will create a new conda environment called with the name defined within the ``environment.ylm`` file. 
 
 To create an ``environment.ylm`` file from a specific conda environment do as follows
 
 ```markdown
->conda activate <ENV NAME>
->conda env export > environment.yml
+conda activate <environment name>
+
+conda env export > environment.yml
 ```
 
 This command will export the list of packages and their versions that are installed in the current active conda environment so that you can quickly share it with anyone.