Google Colab + Kaggle - Downloading Datasets & Uploading Submissions from a Notebook

แชร์
ฝัง
  • เผยแพร่เมื่อ 19 ธ.ค. 2024

ความคิดเห็น •

  • @saswatuna7707
    @saswatuna7707 2 ปีที่แล้ว

    Could you do a video over how to clean data from a dataset in Kaggle

  • @taiyosuzuki2637
    @taiyosuzuki2637 3 ปีที่แล้ว +1

    Hello, Adrian.
    When I download the file, all the files are downloaded randomly.
    How can I get it to download like a kaggle file?
    Is there any way to avoid this?

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      Can you explain a bit more on your issue? Are you having issues with downloading larger datasets with multiple files such as the "A Large Scale Fish" dataset which is 3 GB and contains multiple folders?

    • @taiyosuzuki2637
      @taiyosuzuki2637 3 ปีที่แล้ว +1

      @@AdrianDolinay Adrian, thank you.
      I wrote some code to configure it, and it worked. I think my first code was wrong. Sorry about that.

  • @wizix9877
    @wizix9877 3 ปีที่แล้ว +1

    does it actually unzip the chess file and load csv?

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      Hi! My preferred method is to read in the zipped file into a pandas DataFrame like I did at 8:57.

    • @wizix9877
      @wizix9877 3 ปีที่แล้ว

      @@AdrianDolinay what i meant is loading the zip file directly into set would not give you proper values...i think you have to unzip it first. unless the read takes care of uncompressing it. check the values u got in the video.

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว +1

      Yup, pandas will unzip it. The documentation outlines how read_csv unzips it under the "compression" parameter. Please let me know if you have any additional questions!
      pandas.pydata.org/docs/reference/api/pandas.read_csv.html

  • @zooltechno7065
    @zooltechno7065 ปีที่แล้ว

    after running command " ! kaggle datasets download -d 'name-of-dataset'"
    it is generating
    Error 403: forbidden

  • @misrahmaqboolofficial
    @misrahmaqboolofficial ปีที่แล้ว

    I couldn't download datasets from gaggle
    As i got message cannot create directory

  • @carbon_molecule
    @carbon_molecule 3 ปีที่แล้ว

    Please help me...
    I used the api properly still
    Whenever I download any competition dataset only CSV and few data(i.e. images ) downloads and most of the file doesn't download

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      Which competition?

    • @carbon_molecule
      @carbon_molecule 3 ปีที่แล้ว

      @@AdrianDolinay dog breed identification

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      @SomeOne try the code below
      import zipfile
      !pip install --upgrade --force-reinstall --no-deps kaggle
      !kaggle competitions download -c dog-breed-identification
      with zipfile.ZipFile('/content/dog-breed-identification.zip', 'r') as zip_ref:
      zip_ref.extractall('/content')

    • @carbon_molecule
      @carbon_molecule 3 ปีที่แล้ว

      @@AdrianDolinay but no zip gets downloaded when I use api to download

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      @@carbon_molecule hmm, when I run the above code all the data is downloaded within Google Colab. Are you using a different IDE? The issue is the current Kaggle API is outdated, but once you run the above code it gets updated.

  • @kelijunior7871
    @kelijunior7871 3 ปีที่แล้ว

    How would I upload datasets submission directly from the notebook?

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      At 9:58 I go over how to submit it, let me know if you still have issues!

    • @kelijunior7871
      @kelijunior7871 3 ปีที่แล้ว

      @@AdrianDolinay I'm having problems because I am brand new to machine learning. I do not understand the proper algorithms I need to input into Google Colab. I am working on Supervised Learning now. I have looked and looked for something to take me step by step in setting up the algorithms in Google Colab, and cannot find nothing. Is there a place I can go to and be able to look at supervised algorithms for K-Nearest Neighbors which falls under supervised learning to do my project?

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      @@kelijunior7871 I just released a video on KNN. If you're looking to implement a KNN classification algorithm it should help. You can access the notebook through GitHub as well. Hopefully it helps!
      KNN Vid on TH-cam - th-cam.com/video/l3TP8wickk4/w-d-xo.html&ab_channel=AdrianDolinay

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      @@kelijunior7871 KNN notebook on GitHub, under "Machine Learning" folder - github.com/tudev/Workshops-2020-2021

  • @ravikshdikola6089
    @ravikshdikola6089 3 ปีที่แล้ว

    when i am doing submission it shows me 400 bad request

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      The first thing I suggest is double checking that the competition name is correct. It also may be an issue on Kaggle's end. If you keep having issues submitting with the API I suggest posting your issue on Kaggle's "Product Feedback" forum. Hope this helps!
      www.kaggle.com/product-feedback?sort=most-comments

    • @ravikshdikola6089
      @ravikshdikola6089 3 ปีที่แล้ว

      @@AdrianDolinay i think its an issue related to api because i have checked everything but again giving me the same error.

  • @carbon_molecule
    @carbon_molecule 3 ปีที่แล้ว

    By the way do you know about Cyber security?

    • @AdrianDolinay
      @AdrianDolinay  3 ปีที่แล้ว

      Yup! I'm interested most by cryptography, I'll continue to post videos exploring that topic and how to implement cryptographic applications within Python. I also have broader interest in cybersecurity that I plan to explore down the line

    • @carbon_molecule
      @carbon_molecule 3 ปีที่แล้ว

      @@AdrianDolinay that's nice.... Can you please help me in continuing my journey from a intermediate to advanced

  • @taiyosuzuki2637
    @taiyosuzuki2637 3 ปีที่แล้ว

    It means there is no folder.

  • @ayushnayak6138
    @ayushnayak6138 3 ปีที่แล้ว

    This was helpful thanks