First steps

Installing conda.

For the latest Miniconda installers for Python 3.12, go to anaconda.com/download. The Miniconda installers are on the same page as the Anaconda Distribution installers, past registration. Download the command line installer.

sh ./Miniconda3-latest-MacOSX-arm64.sh

list packages

conda list  # Verifica paquetes de Conda
pip list    # Verifica paquetes de Pip
#
conda list | grep -E "numpy|pandas|torch|numba|implicit"
pip list | grep -E "implicit|umap|llvmlite|implicit"

Update conda channels or repositories

conda update -n base -c defaults conda

Create the environment with the provide config file yml

conda env create -f environment_uoc20242pec4.yml python=3.13.2
conda activate uoc20242pec4

update the environment with the provide config file yml

conda env update -f environment_uoc20242pec4.yml --prune
conda env update -n uoc20242pec4 -f environment_uoc20242pec4.yml --prune
conda activate uoc20242pec4

remove env

conda env remove -n uoc20242pec4
conda clean --all

Activate

conda activate uoc20242pec4

Deactivate

conda deactivate

List environments

conda env list

Clean cache

conda clean --all

Remove environment

conda env remove -n uoc20242pec4

Install kernel for jupyter notebook

Install the jupyter server to execute notebooks

conda install jupyter

After that, and after create environment. Select it in the bottom left corner:

All the import errors disappear from the notebook Setup the notebook environment in the upper right corner

alternative

Open the jupyter notebook. On the top right corner, choose configure Jupyter server.

Select python interpreter, choose the uoc20242pec3

Somtimes is needed to click on reload bottom

Apply the changes

You can do the same on the down right corner

And the end, the jupyter notebook has to be at the images shows, using the uoc20242pec3 environment.

List

jupyter kernelspec list

Install kernel

The notebooks use a kernel. The kernel is the environment which the libraries. When you are using intellij o pycharm, you have to select a kernel or a new kernel for your notebook. In the down-right corner, you can see the kernel for the notebook. In the top-right corner, you can choose the kernel for the notebook.

python -m ipykernel install --user --name=uoc20242pec4

Be sure that the env is the your custom environment created previously

which python
#> /Users/xxx/miniconda3/envs/environment_uoc20242pec4.yml/bin/python

Remove/Uninstall kernel

The notebooks use a kernel. The kernel is the environment which the libraries.

jupyter kernelspec uninstall uoc20242pec4

Install repositories or channels

a channel is the place where the packages are located at. For instance, the channel is the artifactory, or another repository online, local o whatever which contains the packages or lib to download on your machine

conda install --channel <CHANNEL_NAME> <PKG_NAME>

Install dependencies

conda install openpyxl

Install pip dependencies

conda activate env_uoc202402pec2
pip install --force-reinstall \
    implicit==0.7.2 \
    llvmlite==0.44.0 \
    numba==0.61.0 \
    pynndescent==0.5.13 \
    tqdm==4.67.1 \
    umap-learn==0.5.7

One by one

pip install --force-reinstall llvmlite==0.44.0  # Requisito de numba
pip install --force-reinstall numba==0.61.0     # Depende de llvmlite
pip install --force-reinstall pynndescent==0.5.13
pip install --force-reinstall umap-learn==0.5.7
pip install --force-reinstall tqdm==4.67.1
# da error
pip install --force-reinstall implicit==0.7.2

review versions

conda list | grep -E "numpy|pandas|torch|numba|implicit"
pip list | grep -E "implicit|umap|llvmlite|implicit"