Hello everyone !
I want to download the whole CXR8 dataset from link mentioned in assignment lab but I dont know
How to download the CXR8 full dataset csv as from provided link ,like train_csv we used in practice . Is I have to make it myself or is it provided there I need yours assisstance .
Hello everyone !
Thanks a lot for your post ! Welcome to the AI4ML Discourse community.
From your question, I can understand that you are having difficulty loading the dataset. When you press on “Shift + Enter” on the cells of the notebook the cells will automatically download the dataset to the desired environment for you.
^ above is a screenshot from the lab notebook.
If you need a quick intro about how to use the notebook here is a quick intro on Youtube: Jupyter Notebook Tutorial for Beginners with Python - YouTube
However if you are curious about the dataset and want to explore it outside of the notebook which Coursera provides you, you can check the academic research paper which was written specifically about this dataset. https://arxiv.org/pdf/1705.02315.pdf
You can download the dataset onto your desired environment outside of Coursera by following this link. Box
I would like to thank you for asking this question on Discourse and please feel free to ask more questions as you move on with the course.
Thanks @canxkoz for responding to this question
Actually I want to download the whole dataset which I can get by link Provided by you I got there but the problem is that I want the CSv file of format like we used in practice , like every image with all 14 pathalogies which binary classified . but csv in CXR8 is not like that so my question is I have to make it myself or I can get it from CXR8
Thank you for responding with a follow up and clarifying your question, I really appreciate it.
After opening your notebook, click on the Jupyter button on the top left side. This will take you to a higher directory where you can see .py files for scoring and a folder called ‘data’.
That ‘data’ folder contains the folder ‘nih’ and inside it you can see the test.csv which I think is exactly what you are looking for.