Project prepared for Data analysis and visualsiation class at MiM UW 2020. Authors: Anna Witwicka & Anna Warno
The goal of this work was to carefully analyze covid-19 situation in UK. The main focus of this study was not only visualisation of basics satistics but also trying to draw some inferences and find trends.
As a framework we chose plotly Dash. It easily integrates with user interface and allows to perform calculations in real time.
Most of the data was downloaded from UK Office for National Satistics (https://www.ons.gov.uk/), also global data like for example https://ourworldindata.org/coronavirus-source-data were used.
The analysis consists of three parts:
Plots with general statistics like total cases, total cases per milion, total deaths etc. are displayed. Additionaly user can choose any group of countries for comparision. Also the data about the location of covid-19 cases are included. Pie chart and two maps (UK and England - the most infected country) Date sliders are avaliable to visualise plots from past.
Interesting facts form three areas: health, psychology and economy were chosen. Plot of differences of 2020 deaths and average from previous 5 years shows how true threat is Covid-19 for UK. Satisfaction from government and concerns compiled with pandemy growth allows to draw inferences about the people opinions changes. Data about the switching to online services and comparision with Germany gave some information about the consumers behaviours in UK.
Two analysis were conducted. Model for total cases prediction basing on RMSE was prepared (Prophet). User can choose a date from calendar and see the model prediction from this date and original curve. Additionaly special events are provided. It allows to examine if some events had an impact on pandemy growth. The second analysis aiming in finding countries with similiar accoording to specified criterion (egz. total cases per milion) situation at the moment. Time series from last 20 days are extracted, and smoothed by Savitzky–Golay filter (https://en.wikipedia.org/wiki/Savitzky%E2%80%93Golay_filter) in order to remove outliers. Then simple RMSE metrics is used for calculating distances between UK time series and other countries (countries with small number of cases were ommitted).
Steps to run the app:
git clone https://github.com/AniaWitwicka/dav_project_uk
cd dav_project_uk
pip install -r requirements.txt
python appnavbar.py