search

gimseng / 99-ML-Learning-Projects

A list of 99 machine learning projects for anyone interested to learn from coding and building projects
MIT License
580 stars 175 forks source link

Added the Ist project problem & solution #8

Closed aarpit1010 closed 4 years ago

aarpit1010 commented 4 years ago

7

I have added the problem as well as the solution. If there needs to be some changes or anything, please feel free to contact. Any inputs would be appreciated.

gimseng commented 4 years ago

@aarpit1010 Thanks for the PR. I roughly look at it, they look great ! There are just a few things about formatting and styles which will be helpful.

First of all, I converted your exercise description file into markdown. Check out github's markdown guide, it usually looks quite nice rendered on GitHub.

Secondly, I'm going to try to see if I could put 'open in google cola' links in the solution folder's readme.md. If you have some ideas how to do that, go ahead and create readme.md or something like that in the solution folder. It will be nice to describe your solutions a little bit in this file as well.

Thirdly, just a comment that its usually a good practice not to work on the master branch, but to create something like 'dev-titanic' branch, and then push/pull from there. Since we haven't written down the proper guide, we shall tentatively allow what you are doing now. For future purpose, do keep that in mind. And apology for not having the right guide up there yet.

I'll check this out later tomorrow to see if I could implement what I suggests above.

aarpit1010 commented 4 years ago

Thank you for your inputs. I'll keep these things in mind in the future and I'll try to provide the link for google colab after opening the jupyter file in google drive colab notebooks. I have explained some points and steps of my solution in the Jupyter notebook itself but if it still needs some work, please do tell and I'll try if possible.

gimseng commented 4 years ago

@aarpit1010 So I found out what I wanted to do with google colab link. It turns out that you need to put some codes in the beginning of your ipynb. Please take a look at this example of ageron's book. Make the link and then update your fork version, and it'll automatically updated here.

And for a markdown file in the solution, its just a few sentences explaining the various files you have there, what which one does. In times, we might have other people chimming in with other solutions, so a over-all markdown file will be good.

aarpit1010 commented 4 years ago

Sir, if this is a link for the Jupyter notebook "https://www.github.com/aarpit1010/Titanic-Dataset-Analysis/blob/master/Titanic%20Survival%20Analysis.ipynb" , then after signing up in google, we can make just one change in the URL to open the notebook in Google Colab, which is : "https://colab.research.google.com/github/aarpit1010/Titanic-Dataset-Analysis/blob/master/Titanic%20Survival%20Analysis.ipynb"

We just have to replace "github.com/..." to "colab.research.google.com/github/..." and it will open in Google Colab (considering the person opening has a signed in account in google).

gimseng commented 4 years ago

@aarpit1010 Yup, that seems to be the case. I was thinking if you could have that link in your .ipynb or in your markdown description file for the solution folder, you could include the google colab links instead. Its your call, but seems like a useful link to have for learners to have a quick way to run the .ipnyb file. Thanks for doing this !

aarpit1010 commented 4 years ago

Sir, I have added the markdown file in solution folder giving instructions to open the notebook in colab. Please verify it and if there are no more changes to be made then let us merge the PR.

Thank you for bearing with my mistakes :\

gimseng commented 4 years ago

@aarpit1010 Thanks so much for getting the first project in !

I cleaned up a little bit here and there. I eventually added the google colab link inside the Jupyter notebook, as I found this to be useful when I open Jupyter notebook on GitHub.

There might be more work to be done to have more documentations both on the exercise and the solution, but i think we have a good base to start. Maybe others (or both of us) later on can create issues to see how to improve upon this.

Again thanks for the submission !