/Sauti-Africa-Market-Monitoring

Sauti Africa Market Monitoring

Primary LanguageJupyter NotebookMIT LicenseMIT

coding duck MX

Data Scientist // Mathematician

TwitterBlogLinkedIn

Sauti-Africa-Market-Monitoring

Sauti Africa Market Monitoring

The sequence of the sripts are the following:

1.- To verify the connection with the database is working:

v2_verify_conn.py

2.- We need to create the schema of the data base. This will also populate the basic tables, such as contries, currencies, sources.

v2_create_schema_db.py

3.- We collect the data from the original database with:

v2_aws_collect_data.py

We need the help of v2_dictionaries_and_lists.py to correct
some ypos or errors in the original database. If it can't be
done, the logs will be dropped in the error_logs table.

4.- Some basic cleaning like dropping prices at zero, trying to fix prices that seem to be missing a decimal point, or otherwise clearly typos are dropped, and splittin the info into retail and wholesale is being doing here:

 v2_split_bc_drop.py

5.- A table with basis stats and statistic for the AD Fuller stationarity test are dropped in the stats table with:

v2_data_stats.py

6.- A meaninful data analysis is done and dropped into their specific table by the code written by Jing Qian (https://github.com/qianjing2020/):

v2_qc_tables.py

7.- Build the bands for the normal, stress, alert or crisis status of the prices with the script:

v2_bands_construction.py