A config.py file is REQUIRED - place in the root directory of this project. It requires a database username as 'username', a database password as 'password', and a database name as 'database_name'
Team Members
- Halley Ngoc Pham
- Liam Baker
- Mark Stephen
For our project we wanted to find and aggregate data that will allow us to look at Covid 19 vaccinations
and Covid 19 recoveries per country by date. The dataset will also allow us to look at key points during
this time period to acertain the effectiveness of the vaccinations at aiding suppression and recovery from
the virus.
In our project we used three seperate Covid 19 data sets.
- Region_date_metadata.csv
- Region_metadata.csv
- vaccinations.csv
These datasets will give us the ability to check relevant covid-19 data across country and date.
Jupyter Notebook
Pandas
Postgres SQL
SQL Alchemy
When building this database we found that our data had a lot of NaN values in multiple columns. To overcome this we
had to change all NaN values to zero values to upload to SQL. Unnecessary rows were filtered by requiring a key piece
of information (number of vaccinations) that must be greater than zero, with other rows either missing data or being
unnecessary.
We also found that when building a database that flowed and made sense we had to tailor which tables held what data.
In doing this we found that we had to drop some columns from the orginal CSV files. Additionaly, some data included
information about regions, which needed to be aggregated per-country for the purporses of this dataset. Not all countries
included sub-regions, and this was required for effective linking to vaccination records in the future.
In doing this we were able to make a database that holds many points of interesting and relevant data that can be joined
in different ways to build in depth analysis on the roll out of Covid 19 vaccinations and the effect this had on the rate
of recovery.
Please note we have built the database in this way to retain as much data as possible, we found that the CSV data in some
instances had missmatching date ranges so chose not to join CSV data before uploading to SQL.
+ vaccinations.csv: https://www.kaggle.com/datasets/priteshraj10/covid-vaccination-all-countries-data, accessed date: 24 October 2022
+ Region_metadata.csv https://www.kaggle.com/datasets/rohanrao/covid19-forecasting-metadata, accessed date: 24 October 2022
+ Region_date_metadata.csv: https://www.kaggle.com/datasets/rohanrao/covid19-forecasting-metadata accessed date: 24 October 2022
To operate this database please do the following.
- Download git repository including all notebooks, schema and resource files.
- Create a config.py file with you postgresql username, password and database name as noted in the "vaccinations.ipynd"
notebook.
- Create a database in progresql using the same name you chose in the config.py, load the schema.sql into your database
using the query tool.
- Once you have your database and config.py set up correctly you can now run the vaccinations.ipynd notebook, this will build
your database.