example.yaml

# Sample config yaml file, describing the table schema and statistical properties
# of the dataset to be generated

# Path to the output file being written
outputFile: ./example_parquet

# Output format. Currently supported format: parquet
outputFormat: parquet

# Number of rows to generate
rows: 10000

# Table column definition
columns:
  - name: id # Column name (Required)
    type: int # Column type (Required)
    ndv_ratio: 0.1 # Number of distinct value (NDV) ratio. #distinct value = rows & ndv_ratio (Default = 1)
    min: 0 # Value range minimum (Optional)
    max: 100 # Value range maximum (Optional)
    distribution: 
      - type: zipfian # Distribution from which values are sampled, currently support: uniform, zipfian, beta, normal
        alpha: 1.5 # alpha parameter for zipfian distribution
  - name: name
    type: string
    ndv_ratio: 1
    # For categorical data like string, distribution are ignored
        
# Sample distribution parameters. To use a specific distribution, user needs to provide necessary parameters,
# e.g., mean and standard variance for normal distribution, alpha and beta for beta distribution, etc.
distribution1:
    - type: normal
      mean: 0
      std: 1

distribution2:
    - type: zipfian
      alpha: 1.5

distribution3:
    - type: beta
      alpha: 1.5
      beta: 0.5