Official site: https://www.streamlit.io/
It's the easiest way for data scientists and machine learning engineers to create beautiful, good-performing apps in hours.
I just tried it because it looks interesting.
-Official document -I touched Streamlit, which makes it easy to create web applications from Python scripts -Streamlit: Front End for Data Scientists -I made a probability distribution simulation app with Streamlit
I made a video of the moving part.
https://www.youtube.com/watch?v=WFV-cBQARbQ
Central limit theorem that the graph of the average value when rolling several dice is close to the normal distribution when the number of dice is increased. % BF% 83% E6% A5% B5% E9% 99% 90% E5% AE% 9A% E7% 90% 86) demo app.
The source code is published on Gist, so anyone can run it by installing Streamlit.
Installation method
pip streamlit
Execution method
streamlit run https://gist.githubusercontent.com/kokuyokugetter/1a7ec1243f469be03fa4574c05bf5018/raw/a49ca69a2f33a9945602caddd607898e2bf0b0ac/clt_demo_dice.py
In short, it seems that you should feed the source code after streamlit run. Since gist can call the raw data page of the raised code, it can be executed from the link without downloading the source.
It was also a mistake to erase what I wrote while trying various things, so the part of trial and error and the code of the reference site are listed as they are, but please do not worry about the comment part (especially the first half) of the source code.
I can tell from the source that I stopped doing it first with numpy.histogram and streamlit.barchart because the graph was messed up.
I'm not sure if the integer value and numpy.histogram are incompatible or streamlit.barchart is incompatible.
Since you can draw on the graph while changing the variable with the slider, it seems that there will be more effective scenes than changing the value of the variable one by one with jupyter and executing it again.