Ctrl+K

Treatment sim streamlit app

Treatment sim streamlit app#

In this exercise you will incrementally build up a simple streamlit app for the treat_sim package. The app will allow a user to set input parameters using sliders, run the simulation model and view a results summary.

It is recommended that you use basic treat_sim script that has already been provided.

Exercise 1#

Create a a basic streamlit app with a title and information.

  • Create a new file called my_streamlit_app.py; copy and paste in the code from main.py

  • Begin to convert this into a streamlit app

  • Import streamlit

  • Add a streamlit title (e.g. “Treatment Simulation model”)

  • Read in and display the markdown text from resources/overview.md Or create the file yourself. Use the following text:

**A simple simulation model of a urgent care and treatment centre.**
* Change the model parameters and rerun to see the effect on waiting times and utilisation of rooms.

  • Usest.spinner and st.success to provide feedback about the model.

  • Run the app as follows streamlit run my_streamlit_app.py

Basic treat_sim script#

'''
Script to complete a basic run of the model and display a table of
results.

The model is imported from a pypi package 'treat_sim'

Full documentation and source code for `treat_sim` is available as 
* Jupyter Book: https://tommonks.github.io/treatment-centre-sim/
* github: https://github.com/TomMonks/treatment-centre-sim

A conda environment has been provided locally,but the model can be pip installed
`pip install treat_sim==1.0.0`

This will be adapted into a basic streamlit app
'''
from treat_sim.model import Scenario, multiple_replications

# set the variables for the run.
# these are just a subset of the total available for this example...
n_triage = 1
n_exam = 3
n_cubicles_1 = 1

# examination mean
exam_mean = 16.0

# runs
replications = 10

# Setup scenario using supplied variables
args = Scenario(n_triage=n_triage, n_exam=n_exam, n_cubicles_1=n_cubicles_1,
                exam_mean=exam_mean)

# run multiple replications of the model.
results = multiple_replications(args, n_reps=replications)

# print mean results.
print(results.mean().round(1))

00_arrivals                       223.9
01a_triage_wait                    35.3
01b_triage_util                     0.6
02a_registration_wait             107.0
02b_registration_util               0.8
03a_examination_wait               27.5
03b_examination_util                0.8
04a_treatment_wait(non_trauma)    141.6
04b_treatment_util(non_trauma)      0.9
05_total_time(non-trauma)         240.2
06a_trauma_wait                   100.4
06b_trauma_util                     0.8
07a_treatment_wait(trauma)        199.9
07b_treatment_util(trauma)          0.9
08_total_time(trauma)             418.1
09_throughput                     151.5
dtype: float64

Exercise 2:#

Update parameters and run the simulation model via the app

  • Convert the hard coded parameter values into streamlit sliders.

  • Set the script up to that streamlit runs the simulation only when a Run Simulation button is clicked.

  • Using st.table or st.dataframe display the results of the simulation model to a user.

Exercise 3:#

Add in a side bar that display the sliders to a user

  • Create a side bar section in your app and add in the slider.

  • Run your model and check it still works.

Exercise 4:#

Research streamlit pages

  • Read the documentation on streamlit pages.

  • Add in an “about page”.

Contents