# Report Structure

###### tags: `others`

[TOC]

https://www.overleaf.com/8234269335cqjtpgzjsbdy

## Keyword

## Motivation and Objective
- Motivation
- Outbreak in Taiwan in May 2021
- Shortage of vaccine
- Is the strategy of vaccine assignment fair?
- Objective
- for non-CS researcher
- as ref for future pandemics
- analysis of diff vacc strategy
- More general
- catch SARS-CoV-2 characteristic

## Related Work
https://hackmd.io/mKc5bxjKRHuWX5c07tOa4Q
- SARS-CoV-2
- ODE-based methods
- take vaccination into consideration
- region-based?
- 中研院

## Contribution

- anal outbreak in May in Taiwan
- find optimal and feasible vacc strat
- 中研院沒vaccine

## Teamwork

- 1 meeting with Tsai and Tyng-Ruey Chuang at EF Wed, 1 meeting only 我們四個
- deploy and accept jobs each meeting
- Work distribution
- ...

## Materials, Design, Methods
https://hackmd.io/iElb_6CrS3yiudH2QTXneQ

    digraph Omodel {
    node [shape=circle] raw_data, input_data, period_data, population_data
    node [shape=box] preprocessor, Initialization_Module, Simulation_Module, Visualization_Module
    
    raw_data -> preprocessor
    preprocessor -> input_data
    input_data -> Initialization_Module
    Initialization_Module -> population_data
    population_data -> Simulation_Module
    Simulation_Module -> period_data
    period_data -> Visualization_Module
    
    {rank = "same"; raw_data; preprocessor;input_data; Initialization_Module}
    rankdir = "LR"
}

- module architecture(參考中研院怎麼寫)
- Data Prepocessing
- Initialization
- Simulation
- Visualization
- initialization
- 2 time periods per day (daytime and night)
- 4 age groups
- worker flow
- people of age 18-64 located at different tract in daytime and at night
- for each tract
- partition contact group method
- 10 types of contact groups
- initial infectors
- vaccination order
- simulation
- $1-(1-p)^n$ speed up
- parallelization?

## Implementation and Experiments
- github
- unix
- C++, python
- Initialization
- input: csv raw data
- data preprocessing
- population per age
- contact probability per type of contact groups
- list of census tracts in Taiwan
- workerflow
- output: graph.txt
- Simulation
- RBT
- Visualization
- num and ratio of each state for each period, tract, age
- Web Framework: `dash`
- Data visualization(line chart, choropleth...): `ploty`
- Deployment: heroku
- comparison of diff scenerios on Taiwan map
- line chart for diff scenerios
- line chart for diff states

The visualization module is presented on a website. The following are the tools we used.
- Programming Language: `Python`
- Web Framework: `Dash`
- Data visualization(line chart, choropleth...): `Ploty`
- Deployment: `Heroku`

In the visualization module, 3 types of charts are presented.
- Choropleth<br>It shows the number of people in different states(For example, deaths or infections) and census tracts(district, township, city). User can select the data type(For example, number of infections), the age group, and 2 scenarioes they want to compare with. After selecting the above options, the user can drag the timeline below to view the changes at different periods, or even play that.
- Line chart 1<br>By selecting options, user can view data in differnt scenarioes.
- Line chart 2<br>By selecting options, user can view data in differnt states.
## Analysis of Performance and Results
- time
- correctness
- Compare with orther results of related works
- scenerios
- ???
## Conclusion

## Reference

- 各種參數來源