---
title: "App-solutely Fabulous: A Data Scientist's Guide to Choosing Python Web Tools Wisely"
author: "Alan Feder"
format: 
  revealjs:
    theme: beige
    transition: slide
    slide-number: true
    footer: "2024 Government & Public Sector R Conference - Alan Feder"
    embed-resources: true
---


## Who am I?

::: columns
::: {.column width="20%"}
![](https://github.com/AlanFeder/nyr-rag-app/blob/main/AJF_Headshot.jpg?raw=true){fig-alt="Alan Feder Headshot" width="4in"}
:::

::: {.column width="80%"}
-   Alan Feder
-   Staff LLM Data Scientist at Magnifi / TIFIN
- Freelance Consultant
- R Expat ![](https://upload.wikimedia.org/wikipedia/commons/thumb/1/1b/R_logo.svg/1200px-R_logo.svg.png){height="0.5in"}  ![](https://cdn-icons-png.flaticon.com/512/664/664866.png){height="0.5in"}  ![](https://upload.wikimedia.org/wikipedia/commons/thumb/c/c3/Python-logo-notext.svg/1200px-Python-logo-notext.svg.png){height="0.5in"}
- LGM ![](https://upload.wikimedia.org/wikipedia/en/thumb/7/7b/New_York_Mets.svg/1200px-New_York_Mets.svg.png){height="0.5in"}
:::
:::

## Data Scientists in R love Shiny


::: columns 
:::: {.column width="50%"}
- I'm a Data Scientist -- I don't even know HTML, CSS, JavaScript, React
- Am I stuck with copying and pasting my graphs to PowerPoint?
::::

:::: {.column width="50%"}
::::: {.fragment}
![](figs/shiny_logo.png)
:::::
::::
:::

## Python has many more options

::: {.columns}
::: {.column width="33%"}
:::: {.fragment}
![](figs/streamlit_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](figs/gradio_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](figs/shiny_logo.png){height="1.4in" fig-align="center"}
::::

:::

::: {.column width="33%"}
:::: {.fragment}
![](figs/flask_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](figs/django_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](figs/jupyter_logo.png){height="1.4in" fig-align="center"}
::::

:::

::: {.column width="33%"}
:::: {.fragment}
![](figs/dash_plotly_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](https://raw.githubusercontent.com/AnswerDotAI/fasthtml/e91b9af04a3e70b0d769f2a4928fdad1f2c67620/nbs/logo.svg){height="1.4in" fig-align="center"}
::::

:::: {.fragment}
![](figs/panel_holoviz_logo.png){height="1.4in" fig-align="center"}
::::
:::
:::

## Too many options

[![](https://imgs.xkcd.com/comics/standards_2x.png)](https://xkcd.com/927/)

## Try (some of) them!


::: {.columns}
:::: {.column width="33%"}
![](figs/streamlit_logo.png){height="1.4in" fig-align="center"}

![](figs/gradio_logo.png){height="1.4in" fig-align="center"}

![](figs/shiny_logo.png){height="1.4in" fig-align="center"}
::::

:::: {.column width="66%"}

- Let's try to make one app with three different methods

- We've already had a lot of different RAG talks this conference - so see how to share your app
 
::::

:::

## Server vs. Serverless


- The computations need to happen _somewhere_ - Python or R needs to be installed somewhere
  - You probably want to share your app with someone, and they need to be able to run it
- If you see "`localhost:8000`", then all the computations are happening locally - so sending the file won't help unless they have all the data and packages
- In-browser versions can often be a good option if you don't have a ton of data
  - BE CAREFUL ABOUT YOUR PASSWORDS AND API KEYS


## Streamlit


::: {.columns}

:::: {.column width="66%"}

- Launched in 2019
- Purchased by Snowflake in 2022 for $800m ![](figs/snowflake_logo.png){height="0.5in"}
- I've been using it a lot for the past 2 years - and even used in a (NY) R Conference talk!
::::

:::: {.column width="33%"}
![](figs/streamlit_name2.png)

::::
:::

## Streamlit - 😃


- Lots of elements to use
  - New ones all the time
- Very Pythonic
  - Simple to transition from script


## Streamlit - 🫤


- Runs top to bottom - not at all reactive
  - Has caching, but it gets confusing very quickly
- Can get very slow if it gets big
- All Streamlit apps kinda look the same
  - Hard to tweak UI




## Streamlit

::: {.columns}
:::: {.column width="75%"}
[![](figs/streamlit_app1.png){height="6in"}](https://dcr-rag.streamlit.app/)
::::

:::: {.column width="25%"}
![](figs/streamlit_host.png)
::::

:::

## Streamlit-Lite



- [Stlite](https://edit.share.stlite.net/)
  - Not officially supported by Streamlit
- Pretty easy interface to copy your Streamlit apps
- I couldn't figure out how to do streaming
- Slower to load, slower to run

- [Link: In-Browser Streamlit](https://www.alanfeder.com/dcr_rag/streamlit_lite.html)


## Gradio


::: {.columns}

:::: {.column width="50%"}

- Started by some Stanford PhD Students in 2019
- Purchased by Hugging Face in December 2021 ![](figs/hf_logo.png){height="0.5in"}
- Powers most Huggingface Spaces
::::

:::: {.column width="50%"}
![](figs/gradio_name.png)

::::
:::

## Gradio - 😃

- Very focused on ML
- Even simpler than Streamlit
- Built in interfaces for Chat
  - Easy to add in features like chat history, Thumbs Up/ Thumbs Down
- Has reactive blocks
- Can turn it into an API


## Gradio - 🫤

- Less flexible than Streamlit
  - Complicated to do multiple things at once
- Fewer components available
  - Sometimes need to use HTML to do what you want
- All Gradio apps look the same -- even more so than Streamlit

## Gradio
 
![](figs/basic_gradio.png){height="6in"}