as.parameters: Converting to parameters-objects

Description Usage Arguments Value Methods (by class) See Also Examples

View source: R/as.parameters.R

Description

Convert vector, matrix, list, or data.frame to parameters-objects.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
as.parameters(..., distr = NULL)

## S3 method for class 'numeric'
as.parameters(..., distr)

## S3 method for class 'matrix'
as.parameters(x, distr, ...)

## S3 method for class 'list'
as.parameters(x, distr, ...)

## S3 method for class 'data.frame'
as.parameters(x, formula, distr, ...)

Arguments

...

parameters of distribution. This can be named vectors or lists, matrices, or data.frames. See examples below.

distr

character giving the distribution. Function of name q\"distr\" has to be available.

x

numeric vector, matrix, list, or data.frame of parameters.

formula

if x is data.frame a formula has to be given.

Value

object of class parameters, see parameters help page.

Methods (by class)

See Also

parameters

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
# Vector input:
as.parameters(loc = 3, scale = 2, shape = .4, distr = "gev")
as.parameters(c(loc = 3, scale = 2, shape = .4), distr = "gev")

# Names can be shortened if unambiguous:
as.parameters(l = 3, sc = 2, sh = .4, distr = "gev")
as.parameters(m = 3, s = 1, distr = "norm")

# Wrong or ambiguous names lead to errors!
## Not run: 
as.parameters(l = 3, s = 2, s = .4, distr = "gev")
as.parameters(loc2 = 3, scale = 2, shape = .4, distr = "gev")

## End(Not run)

# If no names are given, a warning is given and they are guessed for gev, gpd, gum, and ln3.
as.parameters(3, 2, .4, distr = "gev")
as.parameters(c(3, 2, .4), distr = "gev")
## Not run: 
as.parameters(3, 2, .2, .4, distr = "gev") #=> doesn't work

## End(Not run)

# Matrix input:
# Parameters in matrices must have either matching rownames or colnames!
as.parameters(cbind(loc = 10, scale = 4, shape = seq(0, .4, .1)), distr = "gev")
as.parameters(rbind(loc = 10, scale = 4, shape = seq(0, .4, .1)), distr = "ln3")

# If no names are given, a guess is made based on number of rows
# or cols according to distribution (and a warning is given).
as.parameters(matrix(1:9, nr = 3), distr = "gev")
as.parameters(matrix(1:8, nc = 2), distr = "gum")


# The same principles apply for list input and data.frames:

# List input:
as.parameters(list(list(mean = 2, sd = 1), list(mean = 0, sd = 1)), distr = "norm")
as.parameters(list(c(m = 2, s = 1), c(m = 0, s = 1)), distr = "norm")
as.parameters(list(c(loc = 2, scale = 1), c(0, 1)), distr = "gum")
## Not run: 
as.parameters(list(c(loc = 2, scale = 1), c(0, 1, 2)), distr = "gum")

## End(Not run)

# Dataframe input:
xdat <- data.frame(station = c(1, 2), mean = c(2, 0), sd = c(1, 1))
as.parameters(xdat, cbind(mean, sd) ~ station, distr = "norm")
as.parameters(xdat, . ~ station, distr = "norm")
as.parameters(xdat, cbind(mean, sd) ~ ., distr = "norm")

xdat <- data.frame(station = c(1, 2), m = c(2, 0), s = c(1, 1))
as.parameters(xdat, cbind(m, s) ~ station, distr = "norm")
## Not run: 
as.parameters(xdat, cbind(m, s) ~ station, distr = "gev")

## End(Not run)

###

# Results of as.parameters can be used in the normal TLMoments-scheme:
# they can be transfered to quantiles or to TLMoments.

xdat <- data.frame(station = c(1, 2), mean = c(2, 0), sd = c(1, 1))
quantiles(as.parameters(xdat, cbind(mean, sd) ~ ., distr = "norm"), c(.99))

# quantile estimation
p <- as.parameters(loc = 3, scale = 2, shape = .4, distr = "gev")
quantiles(p, c(.9, .95))
p <- as.parameters(cbind(loc = 10, scale = 4, shape = seq(0, .4, .1)), distr = "gev")
quantiles(p, c(.9, .95))
p <- as.parameters(list(list(mean = 2, sd = 1), list(mean = 0, sd = 1)), distr = "norm")
quantiles(p, c(.95, .975))

# With magrittr
library(magrittr)
as.parameters(loc = 3, scale = 2, shape = .4, distr = "gev") %>% quantiles(c(.9, .99))

TLMoments documentation built on Dec. 4, 2019, 5:06 p.m.