Function smartnoise_runtime::components::resize::resize_float

pub fn resize_float(
    data: ArrayD<Float>, 
    resize_config: RowResizeConfig, 
    number_cols: Option<i64>, 
    distribution: &str, 
    lower: ArrayD<Float>, 
    upper: ArrayD<Float>, 
    shift: Option<ArrayD<Float>>, 
    scale: Option<ArrayD<Float>>, 
    privacy_definition: &Option<PrivacyDefinition>
) -> Result<ArrayD<Float>>

Resizes data (made up exclusively of f64) based on estimate of n and true size of data.

Notice that some arguments are denoted with Option<> – this is because not every distribution used for imputation (if necessary) uses every argument (e.g. Uniform does not use shift or scale).

NOTE: If more distributions are added here, their corresponding arguments must be added as inputs.

Arguments

• data - The data to be resized
• n - An estimate of the size of the data – this could be the guess of the user, or the result of a DP release
• distribution - The distribution to be used when imputing records
• lower - A lower bound on data elements
• upper - An upper bound on data elements
• shift - The shift (expectation) argument for the Gaussian distribution
• scale - The scale (standard deviation) argument for the Gaussian distribution
• minimum_rows - The minimum number of rows in the output dataset. Same as number_rows, but will not sample down.
• p - Maximum proportion of dataset included in output

Return

A resized version of data consistent with the provided n