I tried to build a Python Machine Learning Streamlit App in 7 Minutes | Coding Challenge

What's happening guys, welcome to the fifth episode of CodeThat!
In this ep I try to build a machine learning app to for deploying scikit-learn models using the titanic dataset using nothing but Python. In this case we used a XGBoost model, Streamlit, Pandas and Numpy!
Get the Code: github.com/nic...
Oh, and don't forget to connect with me!
LinkedIn: bit.ly/324Epgo
Facebook: bit.ly/3mB1sZD
GitHub: bit.ly/3mDJllD
Patreon: bit.ly/2OCn3UW
Join the Discussion on Discord: bit.ly/3dQiZsV
Happy coding!
Nick
P.s. Let me know how you go and drop a comment if you need a hand!
#machinelearning #codingchallenge #tkintertutorial #python

[@93hothead]

I have to say you are the best machine learning KZbin channel. Other channels always assume the viewer is already at an advanced level.

[@NicholasRenotte]

Thanks @93hothead, trying it make it a little more accessible!!

[@zkiyyeller3525]

YOU ARE AWESOME. Love how light-hearted you are about not making it on time. Inspiring non the less.

[@BeastyPete98]

You have been a life changer, your channel is so much more practical than what cs programs teach

[@sahilverma9330]

Man you content is hella addictive. The videos are always jam-packed with useful info which is presented in such a slick way. Nick, i just got one doubt, i am super noob in ml ai deep learning. i am not sure where to start or what to do. can you just tell us or make a video regarding the step bystep process of how you got to know soooo much about this field and how to attain the same level of knowledge as you do. like what to study to reach this point

[@NicholasRenotte]

You got it!

[@sadra2637]

Just as always, amazing and very nice to watch 👌. Keep it up Nick.😃

[@NicholasRenotte]

Glad you enjoyed it @Sadra!

[@handle_gc]

Streamlit is great. I have built UI for Satellite image processing app in very few lines.

[@NicholasRenotte]

It's great right?!

[@handle_gc]

@@NicholasRenotte Thanks for the ❤️ In case you wanna dig deeper, go for: "Getting Started with Streamlit for Data Science" by Tyler Richards.

[@einsteinsboi]

Another fab tutorial but whyyyyy did you limit yourself to 7 minutes this time LOL? Glutton for punishment ay? The vid edits are hysterical btw... love it! Thanks for sharing your mad skillz with us... also now my inner kiwi has been switched back to down under lingo, will be using using doco and stacko in regular convos... chur mate!! :)

[@luis96xd]

Yeah, that's true!!! Why 7 minutes 🤯

[@NicholasRenotte]

I can't help myself Mary 😂, stoked you're enjoying them!!

[@shubhamprasad6910]

It was a great tutorial on how to make a streamlit application. I was looking for how the xgboost pipeline was made.

[@malikahmed5281]

His way of representation is soo good that why we are here at first second...

[@NicholasRenotte]

Thanks a million @Malik, glad to have you here!!

[@malikahmed5281]

I'm also making a final year project from your idea of Toxic comment...love that.

[@bartoszhojnacki2233]

Again awesome stuff! Also really nice intro to streamlit. Is it possible that one of the next episodes will be NLP related?

[@NicholasRenotte]

Possibly!!

[@polinalee9128]

I noticed that the .fit function inside the utils.py file on line 8 - self.ageImputer.fit(X[['Age']]) - only has a single input, and X. In the tutorials I've seen that usually the line is formatted as 'model.fit(x_train, y_train)'. Why is it different in this case?

[@luis96xd]

Amazing video! Great challenge, now I learned how to use streamlit 😁😁 Thanks

[@NicholasRenotte]

YESS, awesome work!!

[@alisaqib5715]

Tbh ... I don't like time constraints... Just want to learn more no matter how much time will it take ...

[@wasgeht2409]

Bro pls tell me that, whyyy are u soooo good ? hahaha

[@NicholasRenotte]

LOL, thanks for checking it out 😂😂

[@najiibrashiidabdi5014]

Thank you nicholas and i will hope to reach 100k in the future

[@NicholasRenotte]

thanks a mil @Najiib!!

[@ChannelBlogTechIndonesia]

Thank you bro, very inspiring... keep up the enthusiasm for your work...

[@JohnMitchellCalif]

that was super fun! Subscribed

[@tsibulsky4900]

Thank you for the video 👍. You put cabin twice instead of embarkation .

[@hope-kz6tb]

Specialy stable diffusion with anvil

[@NicholasRenotte]

Like an image generator app?

[@hope-kz6tb]

Yes yes

[@hope-kz6tb]

Deploy own stable diffusion on hugging face spaces and use it as backend and then use anvil as frunt end and please also tell how we can put adsense ads code in anvil website

[@hope-kz6tb]

I know its look like i am asking you about creating a complete project but i like your teaching method so it will be helpful for me if you make some videos on this

[@theswissdevelopers4107]

Hey Nicholas, Great video but I cannot run the script. Can you tell me the versions you used for: scikit-learn, streamlit, xgboost? Thank you!

[@MusabShaikh-xq6mv]

Hey Nicolas i have one doubt. How can we deploy a trained GAN model on backend and feed it text input. I tried searching on web but there is no documentation or blog about this

[@motohomemadness]

Hey Nick, thank you so much for sharing this stuff. In the Jupiter NB for this on training the model, after mlfit, you reference model but it doesn't exists yet. What am I missing?

[@rajharshit4365]

yeah same with me! u got any updates?

[@danishsohail6948]

Exactly which utils library is it? bcz its giving me error cannot import PrepProcessor and columns from utils even though i pip install utils

[@hugaexpl0it]

10:14 Bro, you pasted "cabin" on top of "embarked", and it was giving me anxiety for the rest of video. Haha

[@chanin-at-streamlit]

That was a super fun challenge!

[@symsth16]

Nice! Thanks a lot bro!

[@gozdemdndr]

Thank you that was really helpful!

[@gustavojuantorena]

Almost 100k 🚀🚀🚀🚀🚀🚀

[@NicholasRenotte]

YEAHHHH, we're getting there!!

[@CognitiveDissonance1]

thank you for the content

[@NicholasRenotte]

Anytime @Kong, glad you enjoyed it!

[@venkatesh2788]

Please build real time video recommendations project and connect with backend (nodejs)

[@NicholasRenotte]

You got it!!

[@shanno88]

404 popup when opening Git repo?

[@NicholasRenotte]

My bad, repo was private. Give it a crack now @Shanno

[@koji4059]

Would love more tutorial videos

[@mohammedzain9876]

how to deploy a tflite model on streamlit

[@NicholasRenotte]

Let me add this to the challenge list!!

[@ezrachua1317]

doesnt work

[@lohitkattimani7668]

I want to know how many retakes you took 😂😂

[@fairyroot1653]

#code_that_suggestion Generate working Akamai sensor data cookie using trained model of an AI

[@nkorochinaechetam2516]

First to watch, comment and enjoy this beautiful content

[@NicholasRenotte]

Heyyy, thanks so much @Nkoro!!

[@syukrikoko5975]

Nice bro..👍👍

[@IvanSedov-i7f]

Спасибо, это было очень интересно

[@fizipcfx]

im gonna open 5k fake account just to get you to 100k subs. I love your content.

[@NicholasRenotte]

🙏🙏🙏 my guy, I'm happy to have you here man!!

[@hope-kz6tb]

I watch your video where you use anvil with ml please can you make more videos on that topic ?

[@kaushalpowar1531]

It's a great tutorial, learned a lot. I think you accidentally pasted cabin at the place of embarked variable at 10:10.

[@piyushguptaaa]

Please build real time detection model and connect it to flutter Ps. Can i get you mail so that we can connect?

[@NicholasRenotte]

Yo!! Best place to connect is here!!

[@fustigate8933]

Nick you're almost at 100k subs!!! Way to go!!!

[@NicholasRenotte]

We’re closing in!!! Thanks for being here @Fustigate!!

[@chinmaybapat3273]

Hey, when u pass the preds to the model, u actually typed cabin twice😂

[@NicholasRenotte]

Oh crap I did, I have no idea how that didn’t throw errors. Good pickup.

[@History-AI-version]

please change that not intelligent thumbnail. please stop showing that weird surprise look.

[@cedricvillani8502]

There’s a 9 year old video on KZbin about a program called ‘Wizard’ and if you are looking for real unfiltered predictions and querying of a db or whatever, with just Drag and Drop.. then check it out

[@priscarogi2183]

Hi there, I am looking for a Streamlit tutor to help me with my project. I am a biginner and my project is simple, but I have never used strealint before and need a tutor asap . Do you think you could help?

[@alx8439]

Mate, you really need to master multicursor editing mode. It will boost your productivity in tasks like this

[@BigggEagle]

Does it able to adjust the font size of the output?

[@rozaomar2446]

what if we have challenges with terminal 😭

[@gplgomes]

One question: Don't the streamlit run over Jupyter Notebook? I converted the app.py to app.ipynb, but it seems that it don't work.

[@balavikashkandukuri6139]

I tried but it didn't worked. So I used Pycharm ide . And working fine

[@NicholasRenotte]

Haven't used it inside of Jupyter before, didn't see a clear answer here: discuss.streamlit.io/t/working-with-jupyter-notebooks/368/5

[@balavikashkandukuri6139]

@@NicholasRenotte for IBM Watson stt. While using continuous as a parameter. Getting type error. Please help me 😭😭😭😭

[@gplgomes]

Another great job. :-) But the link for the source code isn' t working.

[@NicholasRenotte]

My bad, repo was set to private. Try now!

[@gplgomes]

@@NicholasRenotte OK. Thank you.

[@etterathe]

Hey Nick, I am a beginner in DS and while watching your videos I've got few questions. What a good EDA consists of? (When I thought about that I made a list: missing values, duplicates, outliers, data types, ranges, distributions, is there something else?) As a senior data scientist do you often help juniors at IBM? Which cases are the most problematic for junior data scientists?

[@etterathe]

And following questions to those above: is there any sense in doing huge EDA? Maybe we should use pandas profiller and in few lines of code we have everything we need?

[@NicholasRenotte]

Heya @etterath, that's a great start and what I usually start with. Could add correlation (and others measures of association) and data viz as well!

[@shreyasnair994]

Hey Nicholas can you please post a video about using aws recognition for real time object detection and triggering a raspberry pi when certain objects are detected using aws lambda, I really would like to know if its even possible and if yes how it can be done.

[@CyberVagabonds]

Great!! With an Selectkbests maybe you finish before, with 4 or 5 columns only

[@infarctus05]

I liked a lot. It’s going to help me through my project in school about world cup predictions😅

[@ahmedgailanqasem5652]

Amazing. Thank you

[@rapunzelle714]

Oh, time limit is going down? Looking forward to "Code that" in 5, 3 and 1 minute 😅

[@NicholasRenotte]

LOL > "And today I'm going to do Object Detection in 30 seconds"

[@nikitaandriievskyi3448]

@@NicholasRenotte lmao, and you also should build and train the OD model from scratch in 30 sec too lol

[@rapunzelle714]

Okay - but now it sounds like you kinda did already have a shorts planned for that 🤣

[@NicholasRenotte]

@@nikitaandriievskyi3448 LOL

[@NicholasRenotte]

@@rapunzelle714 👀👀