Undertakes paired comparisons for ranking a set of coefficients, considering whether the differences between the coefficients are significant or not.

The index numbers are based on the model coefficient table that comes straight out of the model, with no sorting.

The function will return a dataframe with the results for each pair of coeeficients tested.

Usage

rankr(
  coeftable = NULL,
  vcovmat = NULL,
  mysongidlist = NULL,
  myaltlookup = NULL
)

Arguments

coeftable: coefficients table from mlogit, with one row per coefficient
vcovmat: variance covariance matrix from mlogit, with one row and one column per coefficient
mysongidlist: a dataframe containing the list of song ids to be tested. It can contain other variables but only songid will be used.
myaltlookup: optional altlookup dataframe (the second element of Repeatr_2()'s return list) used to translate mysongidlist's songid values into coeftable's alt-indexed rows, and to attach song names to the results - songid and alt are different scales (songid spans every classified song, alt only the min_song_count-eligible ones actually fit by the model), so this translation is required, not optional bookkeeping. If omitted the currently lazy-loaded default will be used. Songs in mysongidlist that aren't in altlookup (i.e. below min_song_count) are dropped with a warning, since they have no coefficient to compare.

Value

A data frame with one row per adjacent pair of songs tested, giving title1, title2, their coefficients (mycoef1, mycoef2), the coefficient difference and its z-statistic, p-value and 95% confidence interval (as produced by diffr()).

Details

rankr

Examples

songstobecompared <- dplyr::slice(summary, seq(from=1, to=92, by=10))
mycomparisons <- rankr(coeftable = results_ml_Repeatr4, vcovmat = vcovmat_ml_Repeatr4,
                        mysongidlist = songstobecompared)
#> 
#>  
#> First coefficient: 3.15160717544381 
#>  
#> Second coefficient: 1.1804771823996 
#>  
#> Difference to be tested: 1.97112999304421 
#>  
#> Variance of the first coefficient: 0.0424198483543461 
#>  
#> Variance of the second coefficient: 0.14101846094101 
#>  
#> Covariance of the two coefficients: 0.0397520462780542 
#>  
#> Z-statistic: 6.11414863943603 
#>  
#> P-statistic: 1.99999999902926 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: 1.33924892272679 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 2.60301106336164 
#>  
#> 
#>  
#> First coefficient: 1.1804771823996 
#>  
#> Second coefficient: 2.99125045503584 
#>  
#> Difference to be tested: -1.81077327263624 
#>  
#> Variance of the first coefficient: 0.14101846094101 
#>  
#> Variance of the second coefficient: 0.0402153553683423 
#>  
#> Covariance of the two coefficients: 0.0387413994726352 
#>  
#> Z-statistic: -5.62170294705247 
#>  
#> P-statistic: 1.8908419439589e-08 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: -2.44209720557863 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: -1.17944933969384 
#>  
#> 
#>  
#> First coefficient: 2.99125045503584 
#>  
#> Second coefficient: 1.55497313773894 
#>  
#> Difference to be tested: 1.43627731729689 
#>  
#> Variance of the first coefficient: 0.0402153553683423 
#>  
#> Variance of the second coefficient: 0.0440670936195828 
#>  
#> Covariance of the two coefficients: 0.0379532956637188 
#>  
#> Z-statistic: 15.6936389764983 
#>  
#> P-statistic: 2 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: 1.25689868203469 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 1.61565595255909 
#>  
#> 
#>  
#> First coefficient: 1.55497313773894 
#>  
#> Second coefficient: 1.7709465185483 
#>  
#> Difference to be tested: -0.215973380809354 
#>  
#> Variance of the first coefficient: 0.0440670936195828 
#>  
#> Variance of the second coefficient: 0.0443255740862942 
#>  
#> Covariance of the two coefficients: 0.037049365818188 
#>  
#> Z-statistic: -1.80644319347025 
#>  
#> P-statistic: 0.0708491398455441 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: -0.45030559115312 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 0.0183588295344116 
#>  
#> 
#>  
#> First coefficient: 1.7709465185483 
#>  
#> Second coefficient: 3.43463368962312 
#>  
#> Difference to be tested: -1.66368717107483 
#>  
#> Variance of the first coefficient: 0.0443255740862942 
#>  
#> Variance of the second coefficient: 0.0416408587411747 
#>  
#> Covariance of the two coefficients: 0.0370539702928315 
#>  
#> Z-statistic: -15.2776631703945 
#>  
#> P-statistic: 1.07735208464099e-52 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: -1.87712471181244 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: -1.45024963033721 
#>  
#> 
#>  
#> First coefficient: 3.43463368962312 
#>  
#> Second coefficient: 2.19875828310069 
#>  
#> Difference to be tested: 1.23587540652243 
#>  
#> Variance of the first coefficient: 0.0416408587411747 
#>  
#> Variance of the second coefficient: 0.0450744117590485 
#>  
#> Covariance of the two coefficients: 0.0379531307884917 
#>  
#> Z-statistic: 11.8872597418323 
#>  
#> P-statistic: 2 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: 1.03210129462517 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 1.43964951841969 
#>  
#> 
#>  
#> First coefficient: 2.19875828310069 
#>  
#> Second coefficient: 2.45670420855044 
#>  
#> Difference to be tested: -0.257945925449744 
#>  
#> Variance of the first coefficient: 0.0450744117590485 
#>  
#> Variance of the second coefficient: 0.0491032660733902 
#>  
#> Covariance of the two coefficients: 0.0385323434380715 
#>  
#> Z-statistic: -1.9718129369467 
#>  
#> P-statistic: 0.0486309649554755 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: -0.514346522285449 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: -0.00154532861403855 
#>  
#> 
#>  
#> First coefficient: 2.45670420855044 
#>  
#> Second coefficient: 2.61508600037117 
#>  
#> Difference to be tested: -0.158381791820735 
#>  
#> Variance of the first coefficient: 0.0491032660733902 
#>  
#> Variance of the second coefficient: 0.0735540408379587 
#>  
#> Covariance of the two coefficients: 0.041617793563402 
#>  
#> Z-statistic: -0.797696099838669 
#>  
#> P-statistic: 0.425046867706681 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: -0.547537902815163 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 0.230774319173694 
#>  
#> 
#>  
#> First coefficient: 2.61508600037117 
#>  
#> Second coefficient: -1.02160744461325 
#>  
#> Difference to be tested: 3.63669344498442 
#>  
#> Variance of the first coefficient: 0.0735540408379587 
#>  
#> Variance of the second coefficient: 0.0997223633400827 
#>  
#> Covariance of the two coefficients: 0.0317249024098666 
#>  
#> Z-statistic: 10.9736960151414 
#>  
#> P-statistic: 2 
#>  
#> Lower boundary of 95% confidence interval of the difference between the two coefficients: 2.9871475543078 
#>  
#> Upper boundary of 95% confidence interval of the difference between the two coefficients: 4.28623933566103 
#>  
#> 
#> Joining with `by = join_by(alt1)`
#> Joining with `by = join_by(alt2)`
mycomparisons
#>                title1              title2  mycoef1   mycoef2 mycoefdiff
#> 1        waiting room         provisional 3.151607  1.180477  1.9711300
#> 2         provisional           blueprint 1.180477  2.991250 -1.8107733
#> 3           blueprint              stacks 2.991250  1.554973  1.4362773
#> 4              stacks returning the screw 1.554973  1.770947 -0.2159734
#> 5 returning the screw      do you like me 1.770947  3.434634 -1.6636872
#> 6      do you like me        back to base 3.434634  2.198758  1.2358754
#> 7        back to base        floating boy 2.198758  2.456704 -0.2579459
#> 8        floating boy            the kill 2.456704  2.615086 -0.1583818
#> 9            the kill  turn off your guns 2.615086 -1.021607  3.6366934
#>           myz          myp  lower95ci    upper95ci
#> 1   6.1141486 2.000000e+00  1.3392489  2.603011063
#> 2  -5.6217029 1.890842e-08 -2.4420972 -1.179449340
#> 3  15.6936390 2.000000e+00  1.2568987  1.615655953
#> 4  -1.8064432 7.084914e-02 -0.4503056  0.018358830
#> 5 -15.2776632 1.077352e-52 -1.8771247 -1.450249630
#> 6  11.8872597 2.000000e+00  1.0321013  1.439649518
#> 7  -1.9718129 4.863096e-02 -0.5143465 -0.001545329
#> 8  -0.7976961 4.250469e-01 -0.5475379  0.230774319
#> 9  10.9736960 2.000000e+00  2.9871476  4.286239336