getDensity: Compute probability density estimate from numerical fields
In avirkki/synergetr: Synthetic Data Generation

One-dimensional numerical kernel density estimation based on density. This function abstacts away database operations and provides sensible defauts for trimming out outlier data, handling computational cost and handling special formats (like Date fields).

1
2
3

getDensity(tbl, cols, con = options("synergetr_con")[[1]], trim = 0.001,
  adj = 1, discrete = FALSE, n = 500,
  sample_max = options("synergetr_sample_max")[[1]], is_date = FALSE)

`tbl`	Table name
`cols`	Column name
`con`	Connection object to the database, or NULL for in-memory data.frames
`trim`	The quentile of data to be discarded from both ends (min and max) for robust estimation of the probability density
`adj`	Kernel width parameter to be passed to `density`
`discrete`	TRUE for integer-valued output
`n`	Default number of point for the kernel density estimation
`sample_max`	Maximum number of data points to read from the source table. Can be used to limit the data transfer from a database (and the consecutive memory consumption and computational costs)
`is_date`	Convert data first into Date before casting it into integers (where day 0 equals 1970-01-01)