IMF - International Monetary Fund

• As a system
• I want to acquire additional datasets from IMF
• in order to diffuse them on DBnomics.

Tasks

• emulate login (MUST)
• emulate bulk download triggering for a given dataset (MUST)
• monitor mailbox to obtain the download link (MUST)
• download 11 datasets (MUST)
• convert datasets to DBnomics JSON (9 with similar CSV format + metadata + 2 Excel files) (MUST)
• acquire metadata for 11 datasets (MUST)
• record the hierarchy of 11 datasets (MUST)
• decide and implement update policy (SHOULD)

Datasets

• WEO: API or bulk dowload (MUST)
• IFS: API or bulk dowload (MUST)
• BOPS: API or bulk dowload (MUST)
• CDIS (new): API or bulk dowload (MUST)
• CPIS: API or bulk dowload (MUST)
• IRFCL (new): API or bulk dowload (MUST)
• DOTS: API or bulk dowload (MUST)
• GFS: API or bulk dowload (MUST)
• FSIs: API or bulk dowload (MUST)
• Primary commodity prices: API or Excel file (MUST)
• Primary commodity prices forecast: API or Excel file (MUST)
• MONA: not timeseries (WONT)
• Goblal Housing Watch: not a dataset (WONT)

Acceptance criteria

• login must be recognized by IMF server
• HTTP requests for downloading a dataset must be identified
• function checking the mailbox must return either the URL for downloading the dataset or a flag indicating that the dataset isn't ready yet. Be careful about messages concerning another dataset.
• the download procedure must acquire 11 datasets and unzip a data file and a metadata file for each of them
• there must be as many converted series to DBnomics Json format as rows in the CSV datafiles (minus 1 for headers)
• dimensions must match the information in the metadata file
• data hierarchy for IMF must contain 11 datasets at the same level
• the scheduling system must record update policy for each of the 11 datasets

Resources

See fetcher descritpion IMF

Technical steps