Check the Predict Method for TVGEV Objects

Check the predict method.

check_predict(
  object,
  ref,
  level = 0.95,
  which = "confint",
  newdate = as.Date("2020-02-01"),
  confintMethod = "proflik"
)

Arguments

object: A TVGEV package.
ref: An object with class "gev.fit" created by the function gev.fit of the ismev package.
level: Confidence level.
which: Not used yet.
newdate: See predict.TVGEV.
confintMethod: See predict.TVGEV.

Value

Nothing. For each return period computed by predict.TVGEV, the plot built by ismev::gev.prof is shown. Two vertical lines are added, showing the confidence limits found by predict.TVGEV with confintMethod = "proflik". Each vertical line should cut the profiled log-likelihood curve at one of its two intersections with the horizontal line.

Details

The comparison with the ismev package is only completed for stationary models because it is difficult otherwise to get profile-likelihood confidence intevals on Return Levels.

Note

ismev::gev.prof does not return the confidence limits which must be evaluated by eye from the displayed graph.

Examples

data(portpirie)
df <- portpirie
df <- within(df, date <- as.Date(sprintf("%4d-01-01", Year)))
fitNSGEV <- try(TVGEV(data = df,
                       response = "SeaLevel", date = "date"))
fitismev <- gev.fit(xdat = df$SeaLev)
#> $conv
#> [1] 0
#> 
#> $nllh
#> [1] -4.339058
#> 
#> $mle
#> [1]  3.87474692  0.19804120 -0.05008773
#> 
#> $se
#> [1] 0.02793211 0.02024610 0.09825633
#> 
check_predict(fitNSGEV, ref = fitismev)
#> 
#> o Finding CI for Return Levels
#> 
#> **************
#>  Lower bounds 
#> **************
#> 
#> Confidence Level:  0.95
#> 
#>   o Date: 2020-02-01
#> 
#>     - Period:   1000
#>         Optimisation status: 3
#>         Iterations: 27
#>         Objective:    4.66
#>         Constraint check -0.0000028
#>         Gradient directions:  0.0011
#> 
#>     - Period:    500
#>         Optimisation status: 3
#>         Iterations: 14
#>         Objective:    4.62
#>         Constraint check -0.0000055
#>         Gradient directions:  0.0014
#> 
#>     - Period:    200
#>         Optimisation status: 3
#>         Iterations: 13
#>         Objective:    4.55
#>         Constraint check -0.0000040
#>         Gradient directions:  0.0050
#> 
#>     - Period:    150
#>         Optimisation status: 3
#>         Iterations: 13
#>         Objective:    4.53
#>         Constraint check -0.0000070
#>         Gradient directions:  0.0022
#> 
#>     - Period:    100
#>         Optimisation status: 3
#>         Iterations: 13
#>         Objective:    4.49
#>         Constraint check -0.0000048
#>         Gradient directions:  0.0032
#> 
#>     - Period:     50
#>         Optimisation status: 3
#>         Iterations: 15
#>         Objective:    4.42
#>         Constraint check -0.0000375
#>         Gradient directions:  0.0010
#> 
#>     - Period:     20
#>         Optimisation status: 3
#>         Iterations: 17
#>         Objective:    4.31
#>         Constraint check -0.0000627
#>         Gradient directions:  0.0093
#> 
#>     - Period:     10
#>         Optimisation status: 3
#>         Iterations: 18
#>         Objective:    4.20
#>         Constraint check -0.0000082
#>         Gradient directions:  0.0014
#> 
#>     - Period:      5
#>         Optimisation status: 3
#>         Iterations: 17
#>         Objective:    4.09
#>         Constraint check -0.0000652
#>         Gradient directions:  0.0030
#> 
#> **************
#>  Upper bounds 
#> **************
#> 
#> Confidence Level:  0.95
#> 
#>   o Date: 2020-02-01
#> 
#>     - Period:   1000
#>         Optimisation status: 3
#>         Iterations: 36
#>         Objective:   -6.47
#>         Constraint check -0.0000006
#>         Gradient directions:  0.0082
#> 
#>     - Period:    500
#>         Optimisation status: 3
#>         Iterations: 11
#>         Objective:   -6.05
#>         Constraint check -0.0000064
#>         Gradient directions:  0.0124
#> 
#>     - Period:    200
#>         Optimisation status: 3
#>         Iterations: 14
#>         Objective:   -5.57
#>         Constraint check -0.0000002
#>         Gradient directions:  0.0032
#> 
#>     - Period:    150
#>         Optimisation status: 3
#>         Iterations: 9
#>         Objective:   -5.44
#>         Constraint check -0.0000052
#>         Gradient directions:  0.0055
#> 
#>     - Period:    100
#>         Optimisation status: 3
#>         Iterations: 11
#>         Objective:   -5.26
#>         Constraint check -0.0000016
#>         Gradient directions:  0.0035
#> 
#>     - Period:     50
#>         Optimisation status: 3
#>         Iterations: 14
#>         Objective:   -4.98
#>         Constraint check -0.0000009
#>         Gradient directions:  0.0059
#> 
#>     - Period:     20
#>         Optimisation status: 3
#>         Iterations: 15
#>         Objective:   -4.66
#>         Constraint check -0.0000354
#>         Gradient directions:  0.0025
#> 
#>     - Period:     10
#>         Optimisation status: 3
#>         Iterations: 20
#>         Objective:   -4.44
#>         Constraint check -0.0000491
#>         Gradient directions:  0.0289
#> 
#>     - Period:      5
#>         Optimisation status: 3
#>         Iterations: 25
#>         Objective:   -4.25
#>         Constraint check -0.0000007
#>         Gradient directions:  0.0037
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval
#> If routine fails, try changing plotting interval

Check the Predict Method for `TVGEV` Objects

Arguments

Value

Details

Note

See also

Examples