when unique user-defined td regressor has usertd as name #54
Description
When we adjust a series in R with {rjd3tramoseats} with more than one regressor. Their names will be printed in the summary :
library("rjd3toolkit")
#>
#> Attaching package: 'rjd3toolkit'
#> The following objects are masked from 'package:stats':
#>
#> aggregate, mad
library("rjd3tramoseats")
y <- austres
regs <- ts(cbind(x = runif(89), y = runif(89)), start = c(1971, 2), frequency = 4)
my_context <- modelling_context(variables = list(td = regs))
spec <- tramoseats_spec("rsa3") |>
set_tradingdays(option = "UserDefined", uservariable = c("td.x", "td.y"))
mod <- tramoseats_fast(y, spec = spec, context = my_context)
summary(mod)
#> Model: TRAMO-SEATS
#> Log-transformation: yes
#> SARIMA model: (0,2,0) (0,1,1)
#>
#> Coefficients
#> Estimate Std. Error T-stat Pr(>|t|)
#> btheta(1) -0.5543 0.1045 -5.306 1.12e-06 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#>
#> Regression model:
#> Estimate Std. Error T-stat Pr(>|t|)
#> td.x -6.203e-06 5.843e-05 -0.106 0.915754
#> td.y 1.777e-05 6.139e-05 0.290 0.773013
#> AO (1974-10-01) 6.749e-04 1.386e-04 4.868 6.34e-06 ***
#> TC (1978-10-01) -7.617e-04 2.075e-04 -3.671 0.000457 ***
#> TC (1981-04-01) -8.711e-04 2.018e-04 -4.317 4.89e-05 ***
#> AO (1988-04-01) -6.199e-04 1.510e-04 -4.106 0.000104 ***
#> LS (1990-07-01) -2.412e-03 2.413e-04 -9.998 2.57e-15 ***
#> LS (1992-01-01) -1.649e-03 2.522e-04 -6.541 7.31e-09 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> Number of observations: 89, Number of effective observations: 83, Number of parameters: 10
#> Loglikelihood: 541.0693, Adjusted loglikelihood: -259.0635
#> Standard error of the regression (ML estimate): 0.000353831
#> AIC: 538.1271, AICc: 541.1826, BIC: 562.3155
#>
#> Decomposition
#> model
#>
#> DIF: 1 -2 1 0 -1 2 -1
#> MA: 1 0 0 0 -0.5542921
#> var: 1
#>
#> trend
#>
#> DIF: 1 -3 3 -1
#> MA: 1 -0.0460761 -0.8879769 0.1580992
#> var: 0.2473462
#>
#> seasonal
#>
#> DIF: 1 1 1 1
#> MA: 1 1.536064 1.149295 0.1042712
#> var: 0.01198268
#>
#> irregular
#>
#> var: 0.0378558
#>
#>
#> Diagnostics
#> Relative contribution of the components to the stationary
#> portion of the variance in the original series,
#> after the removal of the long term trend (in %)
#>
#> Component
#> cycle 62.348
#> seasonal 0.781
#> irregular 0.030
#> calendar 0.001
#> others 28.397
#> total 91.557
#>
#> Residual seasonality tests
#> P.value
#> seas.ftest.i 0.879
#> seas.ftest.sa 0.989
#> seas.qstest.i 1.000
#> seas.qstest.sa 1.000
#> td.ftest.i 0.271
#> td.ftest.sa 0.993
#>
#>
#> Final
#> Last values
#> series sa trend seas irr
#> 1992 Q3 17526.0 17529.03 17528.43 0.9998270 1.0000346
#> 1992 Q4 17568.7 17571.31 17572.37 0.9998513 0.9999400
#> 1993 Q1 17627.1 17619.09 17618.39 1.0004547 1.0000398
#> 1993 Q2 17661.5 17663.32 17663.45 0.9998970 0.9999926Created on 2024-12-18 with reprex v2.1.1
On the other hand, if we only use one regressor, the printed name will be usertd which is not very indicative:
spec <- tramoseats_spec("rsa3") |>
set_tradingdays(option = "UserDefined", uservariable = "td.x")
mod <- tramoseats_fast(y, spec = spec, context = my_context)
summary(mod)
#> Model: TRAMO-SEATS
#> Log-transformation: yes
#> SARIMA model: (1,1,0) (0,1,1)
#>
#> Coefficients
#> Estimate Std. Error T-stat Pr(>|t|)
#> phi(1) -0.81780 0.07336 -11.148 < 2e-16 ***
#> btheta(1) -0.69530 0.09943 -6.993 8.13e-10 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#>
#> Regression model:
#> Estimate Std. Error T-stat Pr(>|t|)
#> usertd 2.646e-05 7.208e-05 0.367 0.714570
#> AO (1974-10-01) 7.232e-04 1.888e-04 3.831 0.000258 ***
#> LS (1990-07-01) -2.496e-03 3.237e-04 -7.712 3.64e-11 ***
#> LS (1992-01-01) -1.557e-03 3.400e-04 -4.579 1.76e-05 ***
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> Number of observations: 89, Number of effective observations: 84, Number of parameters: 7
#> Loglikelihood: 530.5199, Adjusted loglikelihood: -279.1126
#> Standard error of the regression (ML estimate): 0.0004296577
#> AIC: 572.2251, AICc: 573.6988, BIC: 589.2408
#>
#> Decomposition
#> model
#>
#> AR: 1 -0.8178025
#> DIF: 1 -1 0 0 -1 1
#> MA: 1 0 0 0 -0.6952956
#> var: 1
#>
#> trend
#>
#> AR: 1 -0.8178025
#> DIF: 1 -2 1
#> MA: 1 -0.08783414 -0.9283223 0.1595119
#> var: 0.2823638
#>
#> seasonal
#>
#> DIF: 1 1 1 1
#> MA: 1 1.529853 1.13771 0.09604498
#> var: 0.006703832
#>
#> irregular
#>
#> var: 0.05443102
#>
#>
#> Diagnostics
#> Relative contribution of the components to the stationary
#> portion of the variance in the original series,
#> after the removal of the long term trend (in %)
#>
#> Component
#> cycle 62.272
#> seasonal 0.739
#> irregular 0.091
#> calendar 0.001
#> others 29.278
#> total 92.382
#>
#> Residual seasonality tests
#> P.value
#> seas.ftest.i 0.811
#> seas.ftest.sa 0.992
#> seas.qstest.i 1.000
#> seas.qstest.sa 1.000
#> td.ftest.i 0.939
#> td.ftest.sa 0.940
#>
#>
#> Final
#> Last values
#> series sa trend seas irr
#> 1992 Q3 17526.0 17528.74 17528.05 0.9998437 1.0000392
#> 1992 Q4 17568.7 17570.53 17571.88 0.9998956 0.9999233
#> 1993 Q1 17627.1 17619.56 17618.40 1.0004281 1.0000654
#> 1993 Q2 17661.5 17663.77 17664.19 0.9998714 0.9999765Created on 2024-12-18 with reprex v2.1.1
Is it possible to display the regressors name instead of usertd in the regression table?
Thanks