Building large scale data apps with Taipy | Head-to-Head against Streamlit

Поділитися
Вставка
  • Опубліковано 18 гру 2024

КОМЕНТАРІ • 29

  • @DataSciencewithHarshit
    @DataSciencewithHarshit  4 місяці тому +2

    So, what do you think about Taipy?

    • @narottamsharma4720
      @narottamsharma4720 4 місяці тому

      Amazing

    • @absar66
      @absar66 4 місяці тому +2

      Business License fee is expensive compared to available alternatives

    • @JayaDattasaiKala
      @JayaDattasaiKala 4 місяці тому

      Theres a lot of room for custom widgets to be developed, this will take a lot for taipy to compete with streamlit ; generally streamlited apps gets transformed into either reactified or any popular frontend fw with django or fast api as backend for becoming prod ready
      I have explored many class A frameworks from python, none of them provides a excel /table copy paste directly into browser dataframe editorwidget other than streamlit.

  • @maruthiprithivi
    @maruthiprithivi 4 місяці тому +1

    Very nicely structured content. Loved that you kept it to the point and didn't add fillers to drag out the duration of the video. High production quality!!

  • @kubectlgetpo
    @kubectlgetpo 3 місяці тому +4

    Streamlit is not designed for production. It is exactly fir prototyping, fail fast, iterate fast. Once it sticks, go proper full stack!

    • @heijun149
      @heijun149 Місяць тому

      Well, now I am using streamlit to do an AI chatbot application. Now wish to be designed for production, do you have any suggestions?

  • @hasanlock
    @hasanlock 26 днів тому

    Amazing presentation 🎉

  • @nerdg2
    @nerdg2 2 дні тому

    I tried this dude, this has the same issues with text input that react native had few months ago! It's not that great and it can never be on production as how my experience went

  • @interspacer4277
    @interspacer4277 4 місяці тому +1

    What about vs. Anvil?

  • @pranaamerta8015
    @pranaamerta8015 4 місяці тому +1

    How's this compared to Shiny for Python

  • @GamingCritique1
    @GamingCritique1 4 місяці тому

    Nice! This makes it so much easier

  • @coding_wiz
    @coding_wiz 4 місяці тому

    will try it for one of my dashboards. Very nice tutorial.

  • @vishwastyagi8390
    @vishwastyagi8390 4 місяці тому

    Great ..we’ll try this Thanks

  • @gagansingh3481
    @gagansingh3481 4 місяці тому

    I love streamlit easy to understand without any such hard coding required , client needs output
    Designer app se koi Lena Dena Hota use

    • @DataSciencewithHarshit
      @DataSciencewithHarshit  4 місяці тому

      Agreed, I also love Streamlit. In fact, comparing it with Streamlit does say a lot about Streamlit. But if I think beyond those use cases, that’s when I will look for an alternative.

  • @AhmedIsam
    @AhmedIsam 4 місяці тому

    There is a shameless cheating @ 9:32 what the hell is Decimator object that reduces number of points from 1 million to 500 !!!
    Why there is no equivalent decimation in streamlit?

    • @DataSciencewithHarshit
      @DataSciencewithHarshit  4 місяці тому

      This is already explained in the video, watch till the end. The whole point is the support for such algos within Taipy.

    • @AhmedIsam
      @AhmedIsam 4 місяці тому +1

      @@DataSciencewithHarshit
      Thanks for the reply, but that decimation support is literally a one-liner. That has nothing to do with web framework.
      Here is the equivalent cheat in streamlit:
      ```python
      df = df.iloc[::len(df)//500][:500] # Decimation
      ```
      I tried it out, streamlit is reduced to 1 second delay, still not as fast as taipy.
      You can still prove your point while being fair, don't make streamlit plot 1 million points and decimate the task to 500 points in taipy.

    • @AhmedIsam
      @AhmedIsam 4 місяці тому

      @DataSciencewithHarshit
      I think it just clicked why Decimator object is a thing. My one-line python equivalent still means that there is round-trip to the backend to resove the reaction to user input. While the decimator object wraps a javascript solution that is front-end resolved, so, no frontend-backend-frontend trip is involved, making it faster.
      I will consider this solution for data intensive apps.

  • @irapk79
    @irapk79 4 місяці тому

    NIce, very nice , CRUD possible ?

    • @DataSciencewithHarshit
      @DataSciencewithHarshit  4 місяці тому

      Yes, check their gallery for more examples. You will have to hitch it up with a DB for CRUD.

    • @irapk79
      @irapk79 4 місяці тому

      @@DataSciencewithHarshit Awesome ,thank you

  • @AlonAvramson
    @AlonAvramson 4 місяці тому

    Looks great ! thank you!

  • @AliAlias
    @AliAlias 4 місяці тому +2

    Streamlit more easy to write code

  • @MohanKumar-gj9th
    @MohanKumar-gj9th 4 місяці тому

    Streamlit is free though

  • @KevinKreger
    @KevinKreger 4 місяці тому

    But flask is not for production.