Streamlit's Hidden Gems 💎

Hey awesome folks!
I'm back again with another round of cool Streamlit apps and components. This post is going to be a long one, so please stay patient and read it till the end...! 😊

😎 Awesome Apps Corner

Get ready to be blown away! 🤯

1. Guess the Country 🌍

Play the game of guessing the right Country or Territory within 6 guesses! Test your geography skills here. 😉

🤔 How does it work?

• An image is shown that outlines the mystery Country or Territory and if you guessed it correctly within 6 chances, then you are the winner! 🎉

• Each incorrect guess will reveal some information like distance, direction & proximity of the guess that will help you find the mystery.

• For more, please read the rules provided in the expander 🏁

  

🛠 Developer Gerard Bentley
🕹 Try the App
📌 Source Code

2. Brainstorming Buddy 💡

An app that helps generate ideas on any given topic with the help of GPT-3 (a language model created by OpenAI).
🤔 How does it work?

• Fill in your topic of interest in the text field

• Press Enter and ideas will be generated below

• You can also download the ideas as a .txt file.

  

🛠 Developer Ayoub Nainia
🕹 Try the App
📌 Source Code

3. Links Page 🤩

A tool to store all your personal links in one place, just like your favorites, Link Tree, or Bio Link.

🤔 How does it work?

Follow the steps mentioned in the source code and you will get your very own Streamlit Bio Link ready! 🎉

🛠 Developer Chanin Nantasenamat
🕹 Try the App
📌 Source Code

4. Color-coded writing 🌈

Here's an app that color-codes your text based on sentence length!

🤔 How does it work?

• Add some text in the field provided

• Press Enter and it will trigger the colorful writing based on the varying length of sentences.

🛠 Developer Johannes Rieke
🕹 Try the App
📌 Source Code

5. Constellation Explorer 🛰

An interactive app that computes a transit schedule from the biggest satellite constellations for over any area on Earth!

🤔 How does it work?

Select the time range, any constellation, your location, etc... and the stats will tell you everything.

📌 I found it interesting though and would recommend first researching this topic a little bit before you explore this app.

🛠 Developer Prit Chovatiya
🕹 Try the App
📌 Source Code

6. Confusion matrix 👀

A simple app to understand Machine Learning algorithms and metrics in the form of an interactive storybook. As of now, it covers the confusion matrix algorithm.

🤔 How does it work?
Well, depends on the option you choose at the start. I would recommend going with the second option though. 🤓

🛠 Developer Sebastian Flores
🕹 Try the App
📌 Source Code

7. More Apps ⭐

Some apps just need a special place, so here are they ;)

👑 Streamlit Component: streamlit_chat

You can now add a chatbot interface inside your Streamlit apps using the streamlit_chat component. A huge shoutout to the creators, Yash Pawar and
Yash Vardhan for making it easier to set up the chat interface. 🤩

Without any further do, let's try it out!

🤝 Let's code together

This time, I have come up with an idea to use the component with a purpose instead of writing the same sample code.

⚡ The idea here is to build a bot that shares random jokes. The jokes will be fetched from a public API here.

Now, create a new file, namely, chatbot.py and follow the steps along...

Steps

1. Install and import the required libraries. 📦

    pip install requests streamlit streamlit-chat
   

    import streamlit as st
    from streamlit_chat import message
    import requests
    import json
   

   Libraries usage:

   • requests - a python library to send HTTP requests to the API

   • streamlit & streamlit_chat - for building an interactive chatbot web app

   • json - to retrieve data from API in this format.

2. Bot greetings and user prompts.

    st.header('😆 LOL: The Sarcastic Joke Teller Bot')
   
    message("Welcome human 🙏", avatar_style='big-smile') 
    message("Here's the joke for today!", avatar_style='big-smile')
   
    # Prompts when the user agrees to get more jokes
    user_prompts = ['yes', 'sure', 'why not?', 'one more', 'definitely', 'yeah!']
   

   • message - function used for displaying bot greeting messages.

     • avatar_style - you are free to set the avatar style parameter from the list here.

   • user_prompts - a list for matching user responses.

3. The main recipe that makes the bot work!

   Let's write a function that carries out certain tasks mentioned in the comments below 👇

    def bot_work():
        # Pass the Joke API URL using requests to fetch jokes
        api_endpoint = "https://official-joke-api.appspot.com/random_joke"
        r = requests.get(api_endpoint)
   
        # Decoding the data into json format
        info = r.json()
        data = json.dumps(info) # serializes data
        joke_scope = json.loads(data)  # deserializes the data into str form
   
        setup  = joke_scope['setup']
        punch = joke_scope['punchline']
   
        # Displaying jokes & asking the user if they want more jokes
        message(setup + ' ' + punch, avatar_style='big-smile')
        message("Can I share another joke?", avatar_style='big-smile')
        chat = st.text_input('Chat here! 💬', ' ')
   
        # If user response matches with prompts listed, then
        # more jokes will be shared, else the bot will stop sharing. 
        if chat in user_prompts:
            message(chat, is_user=True, avatar_style='miniavs')
   
        if st.button('No more Jokes!🙅‍♀️'):
            message("Okay! Sayonara 👋", avatar_style='big-smile')   
   
    # Calling the function to test our bot
    bot_work()
   

4. Save all the code and witness the bot live.
   Run the given command in a conda virtual environment and make sure that you're running the file from the saved location.

    streamlit run chatbot.py
   

   If all goes well, then Chrome will open up your app on port localhost:8501 in the browser! 🎉

   

   🕹 Open the Demo App
   📌 Source Code

✨Special Mentions

Some components needed a highlight!

• streamlit-analytics : Just with this line of code, with streamlit_analytics.track() , it can count page views as well as track & visualize all widget interactions across users directly in your browser!

  

  🛠 Creator - Johannes Rieke
  🕹 Try the Demo App
  📌 Source Code

• streamlit-extras : A library of small custom Streamlit components where you can discover, try, share, and install extras using pip install with ease!

  

  🛠 Creator - Arnaud Miribel
  🕹 Try the App
  📌 Source Code

Conclusion

🎪 For more such amazing components, refer: https://components.streamlit.app/

Well, that's all for this post. I will be back with more amazing stuff next time.

If you enjoyed this post, please drop your feedback and thoughts in the comments below, and keep the conversation going!

👋 Connect with me on Twitter, GitHub, and LinkedIn.

Happy Streamlit-ing! 🎈