The goal of validatesuggest is to generate suggestions for validation rules from a supplied dataset. These can be used as a starting point for a rule set and are to be adjusted by domain experts.
And the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("data-cleaning/validatesuggest")library(validate)
library(validatesuggest)
data(retailers, package="validate")
data(SBS2000, package="validate")
# does all
#suggest_rules(retailers)
suggest_pos_check(retailers)
#> Object of class 'validator' with 7 elements:
#> PC1: incl.prob >= 0
#> PC2: staff >= 0
#> PC3: turnover >= 0
#> PC4: total.rev >= 0
#> PC5: staff.costs >= 0
#> PC6: total.costs >= 0
#> PC7: vat >= 0
suggest_range_check(retailers, min=TRUE, max=TRUE)
#> Object of class 'validator' with 10 elements:
#> RC1 : size %in% c("sc0", "sc3", "sc1", "sc2")
#> RC2 : in_range(incl.prob, 0.02, 0.14)
#> RC3 : in_range(staff, 1, 75)
#> RC4 : in_range(turnover, 1, 931397)
#> RC5 : in_range(other.rev, -33, 98350)
#> RC6 : in_range(total.rev, 25, 931397)
#> RC7 : in_range(staff.costs, 2, 221302)
#> RC8 : in_range(total.costs, 22, 2725410)
#> RC9 : in_range(profit, -222, 225493)
#> RC10: in_range(vat, 41, 9655)
suggest_na_check(retailers)
#> Object of class 'validator' with 2 elements:
#> NA1: is.complete(size)
#> NA2: is.complete(incl.prob)
suggest_unique_check(SBS2000)
#> Object of class 'validator' with 1 elements:
#> UN1: all_unique(id)
suggest_type_check(retailers)
#> Object of class 'validator' with 10 elements:
#> TC1 : is.factor(size)
#> TC2 : is.numeric(incl.prob)
#> TC3 : is.integer(staff)
#> TC4 : is.integer(turnover)
#> TC5 : is.integer(other.rev)
#> TC6 : is.integer(total.rev)
#> TC7 : is.integer(staff.costs)
#> TC8 : is.integer(total.costs)
#> TC9 : is.integer(profit)
#> TC10: is.integer(vat)
suggest_ratio_check(retailers)
#> Object of class 'validator' with 10 elements:
#> RA1 : turnover >= 0 * total.rev
#> RA2 : turnover <= 9.07 * total.rev
#> RA3 : other.rev >= -0.1 * staff.costs
#> RA4 : other.rev <= 34.55 * staff.costs
#> RA5 : other.rev >= -0.01 * total.costs
#> RA6 : other.rev <= 1.27 * total.costs
#> RA7 : staff.costs >= 0 * total.costs
#> RA8 : staff.costs <= 0.99 * total.costs
#> RA9 : other.rev >= -2.8 * profit
#> RA10: other.rev <= 4.72 * profit
write_cond_rule(car_owner)
#>
#> # Conditional checks
#> if (driver_license == FALSE) owns_car == FALSE
#> if (owns_car == TRUE) income > 0