ML Models

ML Models for Cortex

Variables

This table describe the required variablas and their uses

Name	Description	Mandatory	Default Value
project_id_src	Source Google Cloud Project: Project where the source data is located which the data models will consume.	Y	N/A
project_id_tgt	Target Google Cloud Project: Project where Data Foundation for SAP predefined data models will be deployed and accessed by end-users. This may or may not be different from the source project.	Y	N/A
dataset_raw_landing	Source BigQuery Dataset: BigQuery dataset where the source SAP data is replicated to or where the test data will be created.	Y	N/A
dataset_cdc_processed	CDC BigQuery Dataset: BigQuery dataset where the CDC processed data lands the latest available records. This may or may not be the same as the source dataset.	Y	N/A
dataset_reporting_tgt	Target BigQuery reporting dataset: BigQuery dataset where the Data Foundation for SAP predefined data models will be deployed.	N	SAP_REPORTING
dataset_models_tgt	Target BigQuery reporting dataset: BigQuery dataset where the Data Foundation for SAP predefined data models will be deployed.	N	SAP_ML_MODELS
mandt	SAP Mandant. Must be 3 character.	Y	800
sql_flavour	Which database target type. Valid values are ECC or S4	N	ECC

Simple local output

If you want to test the output of the jinja template locally you can use jinja-cli for a quick check:

1. First install jinja-cli:

pip install jinja-cli

2. Then create a json file with the required input data:

cat  <<EOF > data.json
  "project_id_src": "your-source-project",
  "project_id_tgt": "your-target-project",
  "dataset_raw_landing": "your-raw-dataset",
  "dataset_cdc_processed": "your-cdc-processed-dataset",
  "dataset_reporting_tgt": "your-reporting-target-dataset-OR-SAP_REPORTING",
  "dataset_models_tgt": "your-mlmodels-target-dataset-OR-ML_MODELS",
  "mandt": "your-mandt-number-800",
  "sql_flavour": "ECC"
}
EOF

Here is what an example looks like

{
  "project_id_src": "kittycorn-dev",
  "project_id_tgt": "kittycorn-dev",
  "dataset_raw_landing": "ECC_REPL",
  "dataset_cdc_processed": "CDC_PROCESSED",
  "dataset_reporting_tgt": "SAP_REPORTING",
  "dataset_models_tgt": "ML_MODELS",
  "mandt": "800",
  "sql_flavour": "ECC"
}

3. Create an output folder

mkdir output

4. Now generate the parsed file:

jinja -d data.json -o ouput/filename.sql filename.sql

Alternatively, if you want to generate all files:

for f in *.sql; do
    echo "processing $f ..."
    jinja -d data.json -o "output/${f}" "${f}"
done

Contributing

All submitions are welcome. Please read our code of conduct and contributions guidelines.