Death-App

Introduction

The Death app was created to analyze data and present the results visually. Data analysis relates to the most common causes of death in countries and different levels of development over the years. You can select the type of analysis and then enter the required input data. As a result, you will receive a graph containing the relevant analysis. You will be able to download each visualization. Application enables you to add your own data.

Requirements

To launch application you need to do poetry installation. To install the defined dependencies for this project, run command below.

poetry install 

Data

Used datasets

To run this application you can use datasets stored on Google Drive. In the initial implementation, the following datasets were used:

1. Death dataset, downloaded from the Our World in Data website - annual-number-of-deaths-by-causes.csv
2. Populations dataset, which contains information on populations by year in countries, downloaded from the World Bank website - population_since_1990.csv
3. HDI countries dataset, which contains data on Human Development Index values by countries and years, downloaded from the Our World in Data website - human-development-index.csv

To download datasets you have to run command below.

poe download

Preprocessing downloaded data to application

After downloading three datasets, you have to preprocess them. Command below prepares new four datasets:

• preprocessed death dataset - death_by_country.csv,
• preprocessed population dataset - population_by_country.csv,
• preprocessed hdi dataset - index_mean_by_country.csv,
• final dataset input_dataset.csv which contains information about death causes and population in each country during specific range of years.

All datasets will be stored in the following folder structure res/data.

To preprocess downloaded datasets you have to run command below.

poe all_preprocess

Your own datasets

Application also allows you to add your own data. You can choose two options:

1. Adding only new death causes data with structure like in annual-number-of-deaths-by-cause.csv
2. Adding two new datasets: death causes dataset and development index dataset

First option

If you have death dataset with structure like dataset downloaded from the Our World in Data website, you have to put it in the following folder structure res/data/raw with name annual-number-of-deaths-by-cause.csv. If you don't have these folders you need to create them in project's root directory. Command below downloads two missing datasets:

• human-development-index.csv from Google Drive
• population_by_country.csv from World Bank Dataset API in defined range of years from death dataset

First dataset will be saved in res/data/raw and the second one in res/data.

To download two missing datasets you have to run command below.

poe download_missing

Preprocessing datasets to application

To preprocess your death dataset and human development index dataset you have to run command below.

poe preprocess

This command prepares new three datasets which will be stored in the following folder structure res/data:

• preprocessed death dataset - death_by_country.csv,
• preprocessed hdi dataset - index_mean_by_country.csv
• final dataset input_dataset.csv which contains information about death causes and population in each country during specific range of years.

Second option

You have to prepare two datasets with specific requirements. You have to put them in the following folder structure res/data/raw with names death_dataset.csv and index_dataset.csv. If you don't have these folders you need to create them in project's root directory.

Death dataset requirements

Columns Name	Type and requirements
country	string, countries name
code	string, countries code
year	integer, year by year;
population	float, 0 >=, in thousands
d: cause of death 1	float, 0 >=, in thousands, name of cause have to start with 'd: '
d: cause of death 2	float, 0 >=, in thousands, name of cause have to start with 'd: '
...	float, 0 >=, in thousands, name of cause have to start with 'd: '
d: cause of death N	float, 0 >=, in thousands, name of cause have to start with 'd: '

Dataset must contain at least one column with cause of death.

Index dataset requirements

Columns Name	Type and requirements
country	string, countries name
code	string, countries code, at least one country code like in death dataset
hdi	float, human development index
development	string, level of development

Dataset must contain at least one row.

To check your data and make this application use it you have to run command below.

poe check 

• If your datasets is correct you will see a message New datasets are correct and application will use your datasets. So you can move to the section Run application.
• If your datasets isn't correct you will see a message with error. You have to change your datasets and try again.

Run application

To run application you have to execute command below.

streamlit run main.py