Newer
Older
EVA RIFA ROVIRA
committed
# WIP
PeriodStandardization <- function(data, data_cor = NULL, accum = NULL, time_dim = 'syear',
EVA RIFA ROVIRA
committed
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
leadtime_dim = 'time', memb_dim = 'ensemble',
ref_period = NULL, cross_validation = FALSE,
handle_infinity = FALSE, param_error = -9999,
method = 'parametric', distribution = 'log-Logistic',
na.rm = FALSE, ncores = NULL) {
# Initial checks
target_dims <- c(leadtime_dim, time_dim, memb_dim)
if (is.null(ref_period)) {
ref.start <- NULL
ref.end <- NULL
} else {
ref.start <- ref_period[[1]]
ref.end <- ref_period[[2]]
}
# Standardization
if (is.null(data_cor)) {
spei <- Apply(data = list(data),
target_dims = target_dims,
fun = .standardization,
leadtime_dim = leadtime_dim,
time_dim = time_dim, memb_dim = memb_dim,
ref_period = ref_period, handle_infinity = handle_infinity,
cross_validation = cross_validation, param_error = param_error,
method = method, distribution = distribution,
na.rm = na.rm, ncores = ncores)$output1
} else {
spei <- Apply(data = list(data, data_cor), target_dims = target_dims,
fun = .standardization,
leadtime_dim = leadtime_dim,
time_dim = time_dim, memb_dim = memb_dim,
ref_period = ref_period, handle_infinity = handle_infinity,
cross_validation = cross_validation, param_error = param_error,
method = method, distribution = distribution,
na.rm = na.rm, ncores = ncores)$output1
}
# add NA
if (!is.null(accum)) {
spei <- Apply(data = list(spei), target_dims = leadtime_dim,
output_dims = leadtime_dim,
fun = function(x, accum, leadtime_dim) {
res <- c(rep(NA, accum-1), x)
return(res)
}, accum = accum, leadtime_dim = leadtime_dim)$output1
}
if (is.null(data_cor)) {
pos <- match(names(dim(data)), names(dim(spei)))
spei <- aperm(spei, pos)
} else {
pos <- match(names(dim(data_cor)), names(dim(spei)))
spei <- aperm(spei, pos)
}
EVA RIFA ROVIRA
committed
return(spei)
}
.standardization <- function(data, data_cor = NULL, leadtime_dim = 'time',
time_dim = 'syear', memb_dim = 'ensemble',
ref_period = NULL, handle_infinity = FALSE,
cross_validation = FALSE, param_error = -9999,
method = 'parametric', distribution = 'log-Logistic',
na.rm = FALSE) {
print(summary(data))
EVA RIFA ROVIRA
committed
# data: [leadtime_dim, time_dim, memb_dim]
# maximum number of parameters needed
nleadtime <- as.numeric(dim(data)[leadtime_dim])
ntime <- as.numeric(dim(data)[time_dim])
nmemb <- as.numeric(dim(data)[memb_dim])
fit = 'ub-pwm'
coef = switch(distribution,
"Gamma" = array(NA, dim = 2, dimnames = list(c('alpha', 'beta'))),
"log-Logistic" = array(NA, dim = 3, dimnames = list(c('xi', 'alpha', 'kappa'))),
"PearsonIII" = array(NA, dim = 3, dimnames = list(c('mu', 'sigma', 'gamma'))))
if (is.null(data_cor)) {
# cross_val = TRUE
spei_mod <- data*NA
params_result <- array(data = NA, dim = c(ntime, nleadtime, length(coef)))
EVA RIFA ROVIRA
committed
for (ff in 1:dim(data)[leadtime_dim]) {
data2 <- data[ff, , ]
dim(data2) <- c(ntime, nmemb)
EVA RIFA ROVIRA
committed
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
if (!is.null(ref.start) && !is.null(ref.end)) {
data.fit <- window(data2, ref.start, ref.end)
} else {
data.fit <- data2
}
for (nsd in 1:dim(data)[time_dim]) {
acu <- as.vector(data.fit[-nsd, ])
acu.sorted <- sort.default(acu, method = "quick")
if (na.rm) {
acu.sorted <- acu.sorted[!is.na(acu.sorted)]
}
if (!any(is.na(acu.sorted))) {
if (length(acu.sorted) != 0) {
acu_sd <- sd(acu.sorted)
if (!is.na(acu_sd) & acu_sd != 0) {
if (distribution != "log-Logistic") {
pze <- sum(acu == 0) / length(acu)
acu.sorted <- acu.sorted[acu.sorted > 0]
}
if (length(acu.sorted) >= 4) {
pwm = switch(fit,
"pp-pwm" = pwm.pp(acu.sorted, -0.35, 0, nmom = 3),
pwm.ub(acu.sorted, nmom = 3)
# TLMoments::PWM(acu.sorted, order = 0:2)
)
lmom <- pwm2lmom(pwm)
if (!(!are.lmom.valid(lmom) || anyNA(lmom[[1]]) || any(is.nan(lmom[[1]])))) {
fortran_vec = c(lmom$lambdas[1:2], lmom$ratios[3])
f_params = switch(distribution,
"log-Logistic" = tryCatch(lmom::pelglo(fortran_vec),
error = function(e){parglo(lmom)$para}),
"Gamma" = tryCatch(lmom::pelgam(fortran_vec),
error = function(e){pargam(lmom)$para}),
"PearsonIII" = tryCatch(lmom::pelpe3(fortran_vec),
error = function(e){parpe3(lmom)$para}))
if (distribution == 'log-Logistic' && fit == 'max-lik') {
f_params = parglo.maxlik(acu.sorted, f_params)$para
}
params_result[nsd, ff, ] <- f_params
EVA RIFA ROVIRA
committed
}
}
if (all(is.na(params_result[nsd, ff, ]))) {
EVA RIFA ROVIRA
committed
cdf_res <- NA
} else {
f_params <- params_result[nsd, ff, ]
EVA RIFA ROVIRA
committed
f_params <- f_params[which(!is.na(f_params))]
cdf_res = switch(distribution,
"log-Logistic" = lmom::cdfglo(data2, f_params),
"Gamma" = lmom::cdfgam(data2, f_params),
"PearsonIII" = lmom::cdfpe3(data2, f_params))
EVA RIFA ROVIRA
committed
}
std_index_cv <- array(qnorm(cdf_res), dim = c(ntime, nmemb))
spei_mod[ff, nsd, ] <- std_index_cv[nsd, ]
}
}
}
}
}
} else {
# cross_val = FALSE
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
spei_mod <- data_cor*NA
params_result <- array(data = NA, dim = c(1, length(coef)))
for (ff in 1:dim(data)[leadtime_dim]) {
data_cor2 <- data_cor[ff, , ]
dim(data_cor2) <- c(1, nmemb)
data2 <- data[ff, , ]
dim(data2) <- c(ntime, nmemb)
if (!is.null(ref.start) && !is.null(ref.end)) {
data.fit <- window(data2, ref.start, ref.end)
} else {
data.fit <- data2
}
acu <- as.vector(data.fit)
acu.sorted <- sort.default(acu, method = "quick")
if (na.rm) {
acu.sorted <- acu.sorted[!is.na(acu.sorted)]
}
if (!any(is.na(acu.sorted))) {
if (length(acu.sorted) != 0) {
acu_sd <- sd(acu.sorted)
if (!is.na(acu_sd) & acu_sd != 0) {
if (distribution != "log-Logistic") {
pze <- sum(acu == 0) / length(acu)
acu.sorted <- acu.sorted[acu.sorted > 0]
}
if (length(acu.sorted) >= 4) {
pwm = switch(fit,
"pp-pwm" = pwm.pp(acu.sorted, -0.35, 0, nmom = 3),
pwm.ub(acu.sorted, nmom = 3)
# TLMoments::PWM(acu.sorted, order = 0:2)
)
lmom <- pwm2lmom(pwm)
if (!(!are.lmom.valid(lmom) || anyNA(lmom[[1]]) || any(is.nan(lmom[[1]])))) {
fortran_vec = c(lmom$lambdas[1:2], lmom$ratios[3])
f_params = switch(distribution,
"log-Logistic" = tryCatch(lmom::pelglo(fortran_vec),
error = function(e){parglo(lmom)$para}),
"Gamma" = tryCatch(lmom::pelgam(fortran_vec),
error = function(e){pargam(lmom)$para}),
"PearsonIII" = tryCatch(lmom::pelpe3(fortran_vec),
error = function(e){parpe3(lmom)$para}))
if (distribution == 'log-Logistic' && fit == 'max-lik') {
f_params = parglo.maxlik(acu.sorted, f_params)$para
}
params_result[1, ] <- f_params
EVA RIFA ROVIRA
committed
}
}
if (all(is.na(params_result[1, ]))) {
cdf_res <- NA
} else {
f_params <- params_result[1, ]
f_params <- f_params[which(!is.na(f_params))]
cdf_res = switch(distribution,
"log-Logistic" = lmom::cdfglo(data_cor2, f_params),
"Gamma" = lmom::cdfgam(data_cor2, f_params),
"PearsonIII" = lmom::cdfpe3(data_cor2, f_params))
}
std_index_cv <- array(qnorm(cdf_res), dim = c(1, nmemb))
spei_mod[ff, , ] <- std_index_cv
}
}
}
}
EVA RIFA ROVIRA
committed
}
return(spei_mod)
}