git for Data Science Made Simple... (Hopefully)

Key Takeaways

hello everyone ken here and in this video i'll explain what git is so that hopefully you can add it into your data science repertoire if you commit to using git it can help you to push your data science projects and career prospects to the next level i'll start with an overview and then go through a process walk through i hope this video pulls you in and by the end you'll understand all of the awful puns that i just made in that intro so what is git at the most basic level git is a version control system think you know track changes in microsoft word get operates locally on your computer and lets us see what the differences are between the files that we've added to folders this allows us to revert to older versions if we make a mistake or it allows us to work on the same files as our team members and not mess anything up people often use git and github interchangeably but they aren't the same thing git is a file structure that sits on your local computer and github is the remote place where we can store the files that we're creating when working with a team it's more common to use github as a main point of reference instead of someone's local computer there are other remote repositories like bitbucket so i thought it was important to differentiate what the difference between these things are now why is git or github important version control is extremely relevant when working on large scale projects especially if you're dealing with a team for individual work git is nice because you can make changes or adjustments to code without worrying about breaking anything you can always revert back or just create a new branch to work on that's independent of the main system as a data scientist we often work with other people git allows us to do this without stepping on each other's toes we can each branch the same folder and work on it at the same time at the end we send it back and can reconcile all the changes that we made this system allows us to do three main things first we track the changes and who makes them second we limit bugs by oversight and third we manage concurrent workflows there's a bonus fourth thing that github allows us to do and that's to document our projects so others can use them in the future it's also nice when you document your projects and employers or prospective employers see them so what does a typical project workflow look like for github first we initialize the folder or clone an existing repository for this we use git init or git clone of a remote path [Music] everything that we want to put in our repository we need to stage staging is just letting the system know that we want to eventually add it to the repo strangely enough we use git add followed by the file name to stage it we can also use git add followed by period to add all of the files in the folder a handy command is git status and that lets you know what has been staged or committed or not [Music] to update our local repository we do git commit this updates the repo with the changes that we'd stage we usually do git commit dash m and then use a message saying what we changed that's just good practice [Music] again all of this is being done locally if you want to upload things to github you need to do some additional steps in this case we usually need to push the changes that we've made this is as simple as doing git push you'll often get some errors so just follow along with what the errors say and you'll be able to work your way out of it you might have to set a remote origin or do something along those lines but reading the error messages here is usually really helpful and really that's the gist of it when you're working by yourself so how do we manage working with different people there are two additional steps when working with a team first you want to make a branch separate from the main one that you'll be working on this way your changes won't impact anyone else needing to work on the code base we do this by doing git check out b and then we make up a branch name you can switch between branches by committing the changes in one and then doing git check out followed by the branch name of your choosing [Music] you do the same process of doing git add git commit and and including a message and then pushing the changes once you've made these changes you'll want to merge them back into the main branch you can do this from the command line but i find it a bit easier to just do it on github this process is called creating a pull request you show which branch you want to merge into the master and you usually tag a few other developers or your manager to review it before the changes are integrated in the pull request you can choose which versions of the code will be used and updated into the main branch after you get a pull request approved it's merged and now becomes part of the master branch account [Music] git and github are so much larger than just this simple walkthrough that i'm talking about but still i think it's important to understand a high level what's going on and why it's useful for data scientists looking to get a job github is the place to showcase your portfolio i've done quite a few project reviews and i recommend documenting your readmes well in work and in your personal projects i also use github for my project from scratch series so you can see what it looks like in practice there as well i hope this video has been informative and that i haven't butchered anything too bad as usual good luck on your data science journey and if you've made it this far into the video i'll bet you're committed to learning you