Description

This is a streamlit app that I created to take data originally from a .csv file, through Underline's Redshift and transformed in dbt to a final production streamlit app for users to interactively interrogate the data

The curated tables that I created as a result of this ELT process are used here to feed the streamlit app. I used pandas, geopandas, pydeck, and shapley to create a streamlit app that displays an interactive map with as many cities as csv files that got added to the S3 bucket. Now, non-technical users can use this app which is internally hosted on Kubernetes to interact with the data.

Here is a quick gif of the working app in motion.

As you can see, the user can add data filters from the left sidebar to specify regions and other business specs, then hover over a specific location point to see relevant tooltip information.

Once the filters are in place, a dataframe below is populated with the selections with information such as addresses and company names so the user (maybe a Sales Rep) can click the button below and download the dataframe for printing and taking it out on the road!

Running the Dockerfile

To create the Dockerfile, you need to run the commands below to build your docker image.

First create a .env file in root

After you pull down the code, create a .env file in the root folder of the repo. Add the following:

REDSHIFT_HOST=aws_location
REDSHIFT_DB=your_db
REDSHIFT_PORT=your_port
REDSHIFT_USER=your_user_name
REDSHIFT_PASS=your_redshift_password

NOTE: dsu package needs adjustment -- REDSHIFT_PORT needs to be cast as int. For now, do this manually until in your file until the package gets corrected.

Build the Docker image

To run this streamlit app, pull down the code and run the following in your terminal:

 $ sudo docker build --network=host -t app . --progress plain

The --progress plain is optional. The sudo is also dependant on your system. I need it on my Ubuntu system, but mac does not.

But the --network=host command is necessary.

This will build the docker image.

Run the docker image

To run the image, run this command (again sudo may not be necessary for you if you are on a mac)

#$ sudo docker run --network=host -p 8501:8501 -t app 
docker run --network=host \
 -p 8501:8501 \
 -v </path/to/your>/.env:/app/.env \
 -t app

You should be able to check your browser for localhost:8501 to see the streamlit app.