If you like this video, watch these next: 👉 Part 2 of this Streamlit (Web Application in Python) series is available here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-8M20LyCZDOY.html 👉 Streamlit playlist of 14 videos available here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-ZZ4B0QUHuNc.html What kind of data science web app would you like to see, would love to hear your thoughts. Comments down below 👇
Thanks Python Engineer, you've got nice tutorial videos and channel, great job! 😃So I am pinning this comment for my subscribers to see your excellent resources. Links to Python Engineer's video on Streamlit is: 🔹ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Klqn--Mu2pE.html
wow this is a great way to deploy projects, I had a hard time trying to present the result without sharing the python notebook or the simple charts, this has made my life way easier.
Yes and No. Yes if each project are significantly different from one another and have extensive library dependencies. No, if the projects share similar libraries. For my projects I usually have 3 different conda environments.
Hey Prof! Just a thought / idea: maybe set up monthly challenges for your viewers, followers, etc. to build projects with the tools and techniques you've shown them and other tools they have. Then invite them to share on your channel (like freecodecamp where I first found you!). It's a win-win for you and they as this can be part of their portfolio project or for experience 'youtubing' or teaching what they know - like you. Just a thought! Maybe gauge your viewers/followers interest on this approach and go for it (if there's enough interest)! God Bless You!
This is awesome!! I deployed a model using flask and it was definitely a tough process.. streamlist seems very straightforward!! Definitely going to use this in the future. Thank you!!👍
Thanks for the comment! It certainly is quicker to work with and deploy at a high-level, will have to explore it some more and will definitely make a follow-up tutorial video on implementing with actual ML model deployment
Hello, i can not install streamlit becouse when i am trying to install it i get the following error message. File "setup.py", line 94, in run self._run_cmake() File "setup.py", line 273, in _run_cmake raise RuntimeError('Not supported on 32-bit Windows') RuntimeError: Not supported on 32-bit Windows Please advice
Hi professor i have executed the same code in jupyter there is no error and everything is fime but i am not getting the web page can u pls help me there.
This video and explanation were awesome beyond words. I and my pet cat watched this video with apt attention. We both loved it very much. Thank you Chanin Sir, thanks to Iqbal sir too
This code can be simplified as follows: from requests.api import head import yfinance as yf import streamlit as st st.write(""" # Simple Stock Price App Shown are the stock closing price and volumne of Google! """ ) data = yf.Ticker('GOOGL')\ .history(period='1d', start='2021-01-01', end='2021-06-16') st.line_chart(data.Close) st.line_chart(data.Volume)
Great video series sir just subscribed to your and channel and I'm gonna feed on your resources. I use Jupyter notebook would any code be different from what you typed??
WOW!! This is my 1st real Python project and it worked!!! I want to change my career so started learning Python and libraries as well as other topics for many months now. Recently I thought about start working on real projects and this is my 1st one. Feel soooo good and confident!! I will follow your other simple projects as well, if there are any.
Glad to hear and glad it helped! Yes, there's plenty more where that came from, I have a playlist of 12 videos so far, more coming up. bit.ly/dataprofessor-streamlit
Hi, the current version of Streamlit 0.71.0 is still supporting the beta_expander and beta_columns functions. If you're using an older version which may not have these functions, you can go ahead and update/re-install the streamlit library.
Your channel is super amazing professor! I've one question, If I have a machine learning model that predicts the results using 14 features. can I build a simple UI that prompts the user for input and then the model returns the results (the prediction results) to the interface using streamlet? if yes, can you please make a full project tutorial similar to this idea?
Hi, I have made the video to show how you can install conda on your computer, check it out here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-sDCtY9Z1bqE.html dp is the name of my conda environment, you can feel free to create a conda environment to any name and when activating the environment you can type "conda activate name-of-environment"
I am usin Jupyter notebook, but I get some error code: streamlit.delta_generator.DeltaGenerator at 0x24e7af9cee0. I don't know why st.line_chart(tickerDf.Close) or st.line_chart(tickerDf.Volume) are not runnig well. Could you help me please?, thank you
Hi, the web app will render properly if used as a script file. To run the web app, type into the command line: streamlit run app.py where app.py is the name of the web app.
can you please make a video on how to make a note-taking app in streamlit I tried to implement it but there are so many problems with my version because while deleting a note (row in the data frame) it messes up because the select_box doesn't get updated and sometimes deletes the row which is not intended to: def Notes(): st.markdown('## Notes') st.markdown('####') df= pd.read_csv('Notes.csv') Seal = st.container() nt,tb=st.columns(2) val=nt.text_area('Edit','Enter Note',height=200) if nt.button('Add note'): if val not in df['All Notes'].tolist() and val!='Enter Note': df.loc[len(df.index)]=val df['.']='.' df.to_csv('Notes.csv',index=False) tb.table(df.rename(columns={'All Notes': 'Highlights'}).set_index('.').sample(2)) with nt.expander('Delete'): delt=st.selectbox('Delete',df['All Notes']) if st.button('Delete'): df= pd.read_csv('Notes.csv') df=df.drop(df[df['All Notes']==delt].index) df.to_csv('Notes.csv',index=False) with Seal.expander('See All'): st.table(df.set_index('.')) Notes() this is my implementation.
Hi, you could try comparing the results from many databases and do a simple consensus, if 3 of 5 gets top scores then those miRNA are deemed important. We have done this in a published paper pubmed.ncbi.nlm.nih.gov/27168985/
Hi Judy, Streamlit is meant to be run as a script (embedded inside a .py file and called via "streamlit run app.py". Yes the results of the classification model can be displayed in Streamlit
can you make a video to run the streamlit application to get the filtered output and the database for streamlit is from a dynamic website using selenium.
Awesome! Glad you can use Python to make the web app with Streamlit. More videos in the Streamlit playlist bit.ly/dataprofessor-streamlit If you decide to take a look in R, consider shiny, it's a great library, I've got a tutorial playlist here bit.ly/dataprofessor-shiny
I loved your tutorial, but I was getting an error when I tried to run it. I already use pipx to install yfinance, but the system tells the error: 'No module named 'yfinance''. How can I fix it?
lol, again a nice way to have things not working by choosing the easyway, currently this simple sample doesnt work, crazylazy coding, the enemy of all solidity
Hey there! I just generated a heatmap in a web app by using the above process. But can you suggest how to download the heatmap as an image? I am stuck on it. I have tried almost everything.
Hi, absolutely, I've made a video in this Streamlit series showing the use of heatmap, please find the video here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-xiBXspqs0dk.html
How can streamlit be integrated with MathJax that sequentially reads data from am Excel file to display mathematically rendered list of formulas in the browser?
I followed the tutorial and run the script from the terminal using streamlit run app.py, but it give me an error NonExistentTimeError: 2010-09-10 00:00:00? Any help.
Perhaps you would like to download the dataset file from the internet? If so, please follow this tutorial how you can download files inside Python likegeeks.com/downloading-files-using-python/
@@DataProfessor I want access data online.. without downloaded in my system (actually I want to access the realtime weather data to predict weather conditions at different places)
@@vickyjaiswal9168 Then you need to hace some sort of access to it through code. There are several ways, easiest is to use website API if it has any, if not you need to write another script that will just download data data you want. You can use BeautifulSoup, Selenium or other web scraping packages. But every case will be different, so for every website you will need new script.
Hi, then you will want to deploy your app to the cloud. To do that, check out these 2 videos: - ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-kXvmqg8hc70.html - ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-zK4Ch6e1zq8.html
I've made some introductory videos here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-wnmDt5ZaJKw.html And a guide to approaching the learning of Python for data science here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-AeUnO1oNv08.html
Can we divert python log messages from loggin module to streamlit textbox component? in other words, instead of printing the log messages in console, can i print that in the stream lit web page ?
Hi! Thanks for this tutorial. Streamlit seems cool and simple. I'm new to python. I've just created an eye blink detector that sends key presses when you blink. Thinking of making this an web app. Will you be able to run an eye blink key presser in the browser in the background and go into other apps/windows trigging key presses? I mean can I make it have system wide impact while running in the background?
Hi, I think the concept of your app would work. You would probably have to use the opencv library so that the webcam will be able to detect the eye blink which would trigger some key presses. As for running in the background, I am not sure about that, would really depend on the internet connection and if there would be any timeout from the hosting server.
Hi, I just had a doubt from the portfolio video you made with Ken. After uploading the files on github ,it only shows the basic html with no css. Can you suggest a solution please?
Hi, have you compiled the HTML and CSS files, you can compile by using the command ./hugo -D (see 38:29 of the DS portfolio video). After compilation look all the HTML files and CSS files are in the "public" folder, you need to upload all contents in here into the GitHub. Then it should work. CSS file is in public/dist/css
@@relil9121 @Data Professor - I am also facing same issue on Windows machine. Ariel Ilagan - Is your issue resolved? If so, Please share the step you have taken. I am using Python 3.x (python is installed directly on windows... no conda env.)
Hello Sir, Thank you for such wonderful tutorial videos. I am facing two issues, let me drop them here. 1. It does not matter how many different smiles I test using the Streamlit app using the same .PKL model i generated, it always predicts the same result. How is that possible? Lets say, if for smiles 1, 2 and 3 the pIC50 is 7.1, 7.2 and 7.3. The same set of values will be predicted for other input smiles. 2. If the descriptor list has 882 then the program will generates 88s results, does not matter how many smiles, i provide.
Thanks for the suggestion. I think that may be possible in the near future, I have made the same suggestion to the developers of Mito (a Python library that adds spreadsheet capability to Jupyter) in a recent podcast last week and the video will be released soon. In the meantime, please check out a video I made on Mito ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-js4iaGQvTAo.html
I needed this so much. I have not been able to showcase my data science skills from about 2 years practising just theories. Thank you so much for this and please keep uploading more projects. Love from India 🇮🇳
Thanks for sharing Simon. Glad it was helpful! I have some project video on using machine learning for bioinformatics project, if you're interested. 🔹Bioinformatics Project from Scratch - Drug Discovery Part 1 (Data Collection and Pre-Processing) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-plVLRashaA8.html 🔹Bioinformatics Project from Scratch - Drug Discovery Part 2 (Exploratory Data Analysis) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-qWVTxfLq2ak.html 🔹Data Science for Computational Drug Discovery using Python ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-VXFFHHoE1wk.html
Thanks a lot professor. I installed and run also.However I am getting only html code in my browser 8501.I saved file .py file from Jupyter nanobook before running it. Could you please let me know what is conda update dp.That is even not working when I tried on my cmd
Hi Srikanti, it is advised to not use Jupyter notebook for making streamlit code, it is preferable to use a IDE such as spyder or a simple text editor such as Atom (which was used in the tutorial, you can download this at atom.io). As for "conda activate dp" that is used to activate a conda environment called dp that is on my computer so if it would not work for you, no worries. Could you retry to code in Atom, let me know if that works.
Thanks for this video...but we can run this through jupyter notebook also by importing library ... in previous webinar with krishna naik.. I have seen u and ur explanation way was real help then I realised that how to learn data science ...
Yes in that you can definitely use Streamlit to build a dashboard, but it's much more than that, you can build data-driven web apps, the imagination is the limit.
Thank you so much for teaching us valuable and quality lessons, please help us with the most useful way and most common methods to use pandas for data wrangling thanks again
Thanks for the question, it’s actually a great question! It will activate a conda environment, which is kind of like a sandbox for the web app that we are building and it is a good practice and perhaps convenient to use environments since if anything goes wrong with the environment (conflicting library versions) then it won’t impact other projects of yours.
I think it may be possible to deploy to hugo with some hacks but streamlit should be easier. I will be releasing a video on deployment of streamlit web app to heroku in the next few days.
subscribed after the live interaction with Krish and Dhaval sir 😍😍😍 I am very enthusiastic about data science and machine learning deep learning your content is great sir thank you
Now every algorithms works in order to show what we want to watch say RU-vid, Google Ads etc.. So sir my question is doesn't it makes us less interested in exploring new things. What if algorithms only shows content based on our past results and never lets us know new things. Suppose i don't have any idea what is deep learning and even didn't hear it before. So how it is likely that one will get to know about it, just by recommendation system or any other search result. Sorry it might sounds weird 😬
This is a great post, not weird at all, very thought provoking! I certainly agree, recommendation systems are designed to keep up hooked to the contents. However, if you logout of your account, especially for youtube, then I start to see more neutral contents that are trending on various topics.
Good Morning sir , i have a collection of 2k images , i have got an assignment to make a website and load the images from a database , I have searched a lot but could not find a reasonable solution. Can you give me any suggestions sir?
The level of difficulty would be similar. They're both intuitive and easier to work with. You can get an idea to product in a short period of time. I will show in future tutorial, please stay tuned.
Hi! @Data Professor, i create some interact variable that i add to my dataset ( for example i have a dataset with 4 columns a,b,c,a*c) and i would like to know how to add this new variable on my dashboard in streamlit. does it exist a way to create it so that when i select a value of a and c for example it multiply those value and asign it to the new variable a*c in the dashboard
Yes, that can be done, you can use input widgets to accept user input and in the backend you can perform the operation (+,-,*, etc.) and display the newly calculated value in the dataframe
Hi, thx for the video. When I am trying to 'streamlit run file_name.py' I get 'Illegal instruction (core dumped)'. Same happens if I run 'streamlit hello' I am using ubuntu (mint 20) and python 3.8.6. Any ideas on that?
Hi, can you try using another computer that has an older version of Python such as 3.7, from prior experience the newest version of Python may have some library compatibility issues.
I'm so appreciated ! by the way, I have some question 1. I have this alarm although I have installed anaconda C:\Users\CTC>conda activate dp Could not find conda environment: dp You can list all discoverable environments with `conda info --envs`. 2. Some softs/program I have have to install - Anaconda, Atom, Python ?
Hi for the first point, dp is the name of the conda environment on my computer so you can replace that with your own, you can create a conda environment by conda create -name any-name-you-want (please see for more details docs.conda.io/projects/conda/en/4.6.1/user-guide/tasks/manage-environments.html)
@@DataProfessor anytime!, made an app which has seaborn heatmaps involved in it. But somehow an error of deprecation due to involvement of pyplot without arguments keeps popping up
