Installation instructions#

We recommend everyone to download and install Anaconda. You are of course free to download and install python and the third party packages in other ways. Note that if you install anaconda, then you will end up installing a lot more packages than you need to. In you don’t want that, then one alternative is to install Miniconda, download the following environment file that contains all the necessary package and install them in a conda environment with the command

conda env create -f environment.yml

This will create a new environment called IN1910, see Virtual environments for more info about virtual environments.

Installing third party packages with conda#

For the course IN1910 you need to install the following software:

This can be installed using the Anaconda distribution which is recommended. There are also other options to install python depending on your operating system. You may also consult the guide from IN1900.

Virtual environments#

The main purpose of a virtual environment is to create an isolated environment where you install all your dependencies. Say that you are working on a project where you need a specific version of a package (say numpy). First thing you might wonder is why would you need a specific version of numpy?. Well, numpy evolves all the time. New functionality is added and some functions might be deprecated and removed. If you are working on a large code base that depends on some old numpy functionality that is removed in the most recent version, then you can either try to upgrade the package to support the most recent version of numpy, or you can just stick to an older version.

Say now, that you are starting on a completely new project, and you want to make use of some brand new functionality in numpy. How do you handle that, without breaking the existing project? This is where virtual environments comes to play.

When using conda, it is easy to have multiple conda environments which are virtual environment within the conda ecosystem. To create a new environment, simply do

conda create --name IN1910

This will ask you to confirm that you want to create a new environment called IN1910. Type y and hit enter. To activate this newly created environment you type

conda activate IN1910

Now you can install a package that you need, e.g

conda install numpy

To deactivate the conda environment you can use the command

conda deactivate

Troubleshoot virtual environments#

Once you have activated the virtual environment using the command

conda activate IN1910

you can verify that you have successfully activated the environment. Do do so you can check which python you are using by executing the command

python -c "import sys; print(sys.executable)"

In my case the result was /Users/henriknf/miniconda3/envs/IN1910/bin/python. You will probably not have the exact same results, but you should at least see IN1910 in the path to python. If you don’t see that then the environment is not properly activated. In this case try to deactivate the conda environment a few times and activate again, i.e

conda deactivate
conda deactivate
conda deactivate
conda activate IN1910

If you still don’t see IN1910 in the path to the python executable, then reach out to one of us or ask fellow student.

Git#

Please consult GitHub docs for a very good guide on how to set up git. You need to first install Git locally on your laptop. Next you need to make sure that you are able to connect to Github using either HTTPS or SSH.

Choosing an editor#

When working with code development you need a proper code editor. Some popular choices includes

For a more extensive list, please checkout the Stackoverflow survey from 2021

C++#

Please see the C++ installation guide