This function allows to define perturbation parameters used to perturb cells in magnitude tables.
ck_params_nums(
type = "top_contr",
top_k = NULL,
ptab,
mult_params,
mu_c = 0,
same_key = TRUE,
use_zero_rkeys = FALSE,
path = NULL
)a character value defining the way to identify the magnifier,
e.g which contributions/values in a cell should be the used in the
perturbation procedure. Possible choices are:
top_contr: the k largest contributions are used. In this case,
it is also required to specify argument top_k
mean: the (weighted) cellmean is used as starting point
range: the difference between largest and smallest contribution
is used.
sum: the (weighted) cellvalue itself is used as starting point
it is ignored if variant is different from top_contr.
Otherwise, a scalar number >0 is expected.
in this argument, one ore more perturbation tables are given as input. the following choices are possible:
an object derived from ptable::create_ptable()
or ptable::create_num_ptable(): this case is the same as specifying a named
list with only a single element "all" (as described below)
a named list where the allowed names are shown below and each element
must be the output of [ptable::create_ptable())] or
ptable::create_num_ptable()
"all": this ptable will be used for all cells; if specified, no
elements named "even" or "odd" must exist.
"even": will be used to look up perturbation values for cells with an
even number of contributors. if specified, also list-element "odd" must
exist.
"odd": will be used to look up perturbation values for cells with an
odd number of contributors; if specified, also list-element "even" must
exist.
"small_cells": if specified, this ptable will be used to extract
perturbation values for very small cells
[ptable::create_ptable())]: R:ptable::create_ptable())
ptable::create_num_ptable(): R:ptable::create_num_ptable()
an object derived with ck_flexparams() or
ck_simpleparams() that contain required parameters for the computation of
the perturbation multiplier
fixed extra protection amount (>= 0) applied to the absolute of
the perturbation value of the first (largest) noise component if the cell is
sensitive. This value defaults to 0 (no additional protection). Please note
that sensitive cells can be defined according using the supp_freq(),
supp_val, supp_p(), supp_nk() and supp_pq() methods. An examples is
given in ?cellkey_pkg.
(logical) should original cell key (TRUE) used for
for finding perturbation values of the largest contributor to a
cell or should a perturbation of the cellkey itself (FALSE) take place.
(logical) scalar defining if record keys of
units not contributing to a specific numeric variables should be
used (TRUE) or ignored (FALSE) when computing cell keys.
a scalar character specifying a path to which the parameters created with this functions should be written to (in yaml format)
an object suitable as input to method $params_nums_set() for the
perturbation of continous variables.
# create a perturbation table using
# functionality from ptable-pkg; see help(pa = "ptable")
# this returns an extra ptable for very small cells
ptab <- ptable::pt_ex_nums(separation = TRUE)
# create parameters
ck_params_nums(
type = "top_contr",
top_k = 3,
ptab = ptab,
mult_params = ck_flexparams(
fp = 1000,
p = c(0.20, 0.03),
epsilon = c(1, 0.5, 0.2),
q = 2),
use_zero_rkeys = TRUE,
mu_c = 3)
#> $params
#> $params$type
#> [1] "top_contr"
#>
#> $params$top_k
#> [1] 3
#>
#> $params$ptab
#> i j p v lb ub type
#> 1: 0 0.0 1.00000000 0.0 0.0000000 1.0000000 all
#> 2: 1 0.0 0.27968425 -1.0 0.0000000 0.2796843 all
#> 3: 1 0.5 0.27968426 -0.5 0.2796843 0.5593685 all
#> 4: 1 1.0 0.27968425 0.0 0.5593685 0.8390528 all
#> 5: 1 1.5 0.05858542 0.5 0.8390528 0.8976382 all
#> ---
#> 216: 5 9.2 0.00001179 4.2 0.9999801 0.9999919 small_cells
#> 217: 5 9.4 0.00000499 4.4 0.9999919 0.9999969 small_cells
#> 218: 5 9.6 0.00000203 4.6 0.9999969 0.9999989 small_cells
#> 219: 5 9.8 0.00000079 4.8 0.9999989 0.9999997 small_cells
#> 220: 5 10.0 0.00000030 5.0 0.9999997 1.0000000 small_cells
#>
#> $params$mu_c
#> [1] 3
#>
#> $params$m_fixed_sq
#> [1] 1
#>
#> $params$zs
#> [1] 4.402255
#>
#> $params$E
#> [1] 1.29
#>
#> $params$mult_params
#> $fp
#> [1] 1000
#>
#> $p_small
#> [1] 0.2
#>
#> $p_large
#> [1] 0.03
#>
#> $epsilon
#> [1] 1.0 0.5 0.2
#>
#> $q
#> [1] 2
#>
#> attr(,"class")
#> [1] "params_m_flex"
#>
#> $params$same_key
#> [1] TRUE
#>
#> $params$use_zero_rkeys
#> [1] TRUE
#>
#> $params$even_odd
#> [1] FALSE
#>
#> $params$separation
#> [1] TRUE
#>
#>
#> $type
#> [1] "params_m_flex"
#>
#> attr(,"class")
#> [1] "ck_params"