Where is the dog breed lesson notebooks from Lesson 3?


(Michael) #1

Where is the dog breed notebook from lesson 3? It is about 40 minutes in and entitled “tmp_lesson1-breeds.ipynb”. I can’t find that file or “lesson1-breeds.ipnyb”


(Navin Kumar) #2

The dog breeds notebook is left as a self assignment which can be done with the help of video and the lesson1 notebook. Since its similar in nature its not provided in the repository.

All the notebooks which begin with tmp-* will not be available to be downloaded to our repository.
.gitignore masks them while cloning…


(Khoo) #3

Hi Navin,
Thank you for the explaination on dog breed

Off topic: Is there any course note on which cell / sequence to run on Lesson 3 Rossman ?
In [53]: columns = [“Date”, “Store”, “Promo”, “StateHoliday”, "SchoolHoliday]
In [31]: df = train[columns]
In [54]: df = test[columns]


(Navin Kumar) #4

I am not clear on what you are asking…
I guess your question is how to run the notebook after the cells you have mentioned… Is there any notes in wiki.

If the above is your query, you can check in Wiki: Lesson 3
. I think the wiki- does not answer your query… kindly check out the wiki if I have missed it…

The rossman notebook is designed to be run interactively… I am refering the below from https://github.com/fastai/fastai/blob/master/courses/dl1/lesson3-rossman.ipynb

1st Iteration: for the training set ie
From cell No: 31 df = train[columns] Exclude Cell No:54
and run until and including Cell No: 49 joined = join_df(joined, df, [‘Store’, ‘Date’])

2nd Iteration . This time for the test set.
from cell No: 53 . columns = [“Date”, “Store”, “Promo”, “StateHoliday”, “SchoolHoliday”]
This time exclude cell No: 31 , Include Cell No:54 df = test[columns]
Exclude Cell No: 49 joined = join_df(joined, df, [‘Store’, ‘Date’])

Henceforth one can proceed executing cell wise till the end of the notebook.

It would be better to understand the flow of the notebook with the help of the video. This post is just a guidance .

hope it helps


(Khoo) #5

Hi Navin,
Yes, thank you for your clarification.
Now I have to learn pandas to figure out how to skip data cleaning, join and df but keep the vars, embdings and data to plug into model.
Cheers!