libraries
#detach("package:remhmm", unload = TRUE)
#devtools::install_github("Mbehbahani/DigiPy/remhmm")
library(remhmm)
library(REHdynamics)
library(remstats)
## Warning: package 'remstats' was built under R version 4.3.2
library(remstimate)
## Warning: package 'remstimate' was built under R version 4.3.2
library("nloptr")
## Warning: package 'nloptr' was built under R version 4.3.2
Parameter settings for Scenario1
m=2 #Number States
q=2 #Number of variables (In REM it means the Number of statistics)
num_obs <- 100
N_relations<-200 #number of relational events in each state
# Hidden State
transition_probs <- matrix(c(0.2,0.8,0.8,0.2), nrow = m,byrow=TRUE)
initial_probs <- c(0.1, 0.9)
# Initial Values for Optimization
gamma0=matrix(c(0.2,0.8,0.8,0.2), nrow = m,byrow=TRUE)
mean0=matrix(c(0.1,0.5,0.2,0.7) , nrow = q,byrow=TRUE)
sd0=matrix(c(0.3,0.4,0.1,0.2), nrow = q,byrow=TRUE)
delta0=c(0.3, 0.7)
vc= matrices_to_vector(gamma0, mean0, sd0, delta0)
Lgamma <- matrix(c(0,0,0,0), nrow = m,byrow=TRUE)
Lmean0 <- rep(-Inf, q * 2)
Lsd0 <- rep(0, q * 2)
Ldelta0 <- rep(0, m)
Lvc <- matrices_to_vector(Lgamma, Lmean0, Lsd0, Ldelta0)
Ugamma <- matrix(c(1,1,1,1), nrow = m,byrow=TRUE)
Umean0 <- rep(Inf, q * 2)
Usd0 <- rep(Inf, q * 2)
Udelta0 <- rep(1, m)
Uvc <- matrices_to_vector(Ugamma, Umean0, Usd0, Udelta0)
lower_bounds <- Lvc
upper_bounds <- Uvc
generate REH data using “generate_rem” function
## code to prepare `covar` dataset
# Generate the actors and their exogenous information for the
# generated datasets
set.seed(27613)
covar <- data.frame(id = 1:10, time = 0, z = rnorm(n = 10))
colnames(covar)[colnames(covar) == "id"] <- "name"
# usethis::use_data(covar,overwrite = TRUE)
## code to prepare `param` dataset
# Normal variable (will be add)
param1 <- list(
"baseline" = -4,
"inertia" = 0.1,
"z_send"=0.2 )
param2 <- list(
"baseline" = -4,
"inertia" = 0.5,
"z_send"=0.7 )
REM data (MainTest(Event effect).Rmd)
# Generate REM of the current state
#num_var<-1
#q<-num_var
#initial_probs
#transition_probs
hidden_states<-numeric(num_obs)
for (t in 1:num_obs) {
if (t == 1) {
# Initial state
hidden_states[t] <- sample(1:m, 1, prob = initial_probs)
} else {
# Transition to the next state
hidden_states[t] <- sample(1:m, 1, prob = transition_probs[hidden_states[t - 1], ])
}
}
preparation to generate datasets
windows defenition
## M<-200 specify M (N_relations at the beginning)
windows <- data.frame(
start = seq(1, by = N_relations, length.out = num_obs),
stop = seq(N_relations, by = N_relations, length.out = num_obs)
)
for calculating the effects. of the statistics
it based on the number of statistics. It should be modified
norm.HMM.mllk(vc, x, m,q)
## [1] -27.88027
Result<-norm.HMM.mle(x, m,q, vc,lower_bounds, upper_bounds)
## Warning in nloptr.add.default.options(opts.user = opts, x0 = x0,
## num_constraints_ineq = num_constraints_ineq, : No termination criterium
## specified, using default (relative x-tolerance = 1e-04)
Result
## $gamma
## [,1] [,2]
## [1,] 0.2612493 0.7387507
## [2,] 0.8529362 0.1470638
##
## $mean
## [,1] [,2]
## [1,] 0.1188562 0.6726295
## [2,] 0.1930384 0.7170707
##
## $sd
## [,1] [,2]
## [1,] 0.07251008 0.45995266
## [2,] 0.05054450 0.07620702
##
## $delta
## [1] 0.6637299 0.3362701
Fstates<-norm.HMM.viterbi(x,Result,m,q)
#hidden_states
#x
cbind(x,hidden_states,Fstates)
## hidden_states Fstates
## [1,] 1.651446765 0.6468862 2 2
## [2,] 0.067023823 0.1972435 1 1
## [3,] 0.319174207 0.7596067 2 2
## [4,] 0.048912816 0.1647565 1 1
## [5,] 0.924011489 0.7201511 2 2
## [6,] 0.091345815 0.1934655 1 1
## [7,] 0.211713958 0.7620410 2 2
## [8,] 0.745664162 0.7376323 2 2
## [9,] 0.185323363 0.7085514 2 2
## [10,] 0.099510572 0.2191988 1 1
## [11,] 0.408163629 0.6691445 2 2
## [12,] 0.066899457 0.2184403 1 1
## [13,] 0.543791479 0.6592160 2 2
## [14,] 0.112607332 0.1905198 1 1
## [15,] 0.365665189 0.6492272 2 2
## [16,] 0.188430071 0.1856045 1 1
## [17,] 0.265863296 0.7044724 2 2
## [18,] 0.217796095 0.2333650 1 1
## [19,] 0.735315212 0.7218976 2 2
## [20,] 0.157304619 0.1749077 1 1
## [21,] 0.675010509 0.7140220 2 2
## [22,] 0.560186432 0.7274596 2 2
## [23,] 0.200348712 0.7764138 2 2
## [24,] 0.215558690 0.1814441 1 1
## [25,] 0.107497035 0.1831628 1 1
## [26,] 0.876443939 0.6305120 2 2
## [27,] 1.159033112 0.6437165 2 2
## [28,] 0.111096487 0.1823991 1 1
## [29,] 0.025122196 0.1583019 1 1
## [30,] 0.151813812 0.1911574 1 1
## [31,] 0.305499622 0.7364174 2 2
## [32,] 0.149530812 0.1984807 1 1
## [33,] 0.479399197 0.6530749 2 2
## [34,] 0.153098616 0.1379333 1 1
## [35,] 0.587742011 0.6889346 2 2
## [36,] 0.013837892 0.2061985 1 1
## [37,] 0.435279015 0.7304410 2 2
## [38,] 0.876162853 0.6683150 2 2
## [39,] 0.894991296 0.6558346 2 2
## [40,] 0.129467720 0.1947648 1 1
## [41,] 0.458407402 0.7549627 2 2
## [42,] 0.154056289 0.7385279 2 2
## [43,] 0.800374890 0.6651944 2 2
## [44,] 0.061732999 0.2054986 1 1
## [45,] 0.941808448 0.6620029 2 2
## [46,] 0.127695989 0.1951152 1 1
## [47,] 0.272353284 0.7658479 2 2
## [48,] 0.196185757 0.0760059 1 1
## [49,] -0.512827620 0.9554744 2 2
## [50,] -0.035187805 0.1685775 1 1
## [51,] 0.066545230 0.7621485 2 2
## [52,] 0.213583740 0.1819388 1 1
## [53,] 0.435218605 0.6892471 2 2
## [54,] 0.008514712 0.2211902 1 1
## [55,] 0.463518037 0.7148787 2 2
## [56,] 0.025729170 0.2373265 1 1
## [57,] -0.060756640 0.8185260 2 2
## [58,] 0.023069327 0.2188862 1 1
## [59,] 0.283285534 0.7641513 2 2
## [60,] 0.549345002 0.6585590 2 2
## [61,] -0.044263139 0.2159724 1 1
## [62,] 0.132744594 0.7737801 2 2
## [63,] 0.134524922 0.1939455 1 1
## [64,] 0.657318713 0.6709126 2 2
## [65,] 0.038003201 0.1943513 1 1
## [66,] 0.116898605 0.7540525 2 2
## [67,] 0.139102490 0.1930833 1 1
## [68,] 0.286346673 0.7783457 2 2
## [69,] 0.817792361 0.6986286 2 2
## [70,] 1.142948316 0.6069873 2 2
## [71,] 0.060949475 0.1901416 1 1
## [72,] 1.040705201 0.6328036 2 2
## [73,] 0.126509346 0.1799825 1 1
## [74,] 0.812668219 0.6540862 2 2
## [75,] -0.018746825 0.2033016 1 1
## [76,] 0.224692652 0.8239120 2 2
## [77,] 0.048125334 0.2214900 1 1
## [78,] 0.562399158 0.6730493 2 2
## [79,] 0.780090569 0.6483673 2 2
## [80,] 0.734276627 0.6720325 2 2
## [81,] 0.996044785 0.6569215 2 2
## [82,] 0.150767675 0.1587435 1 1
## [83,] 0.622330613 0.6423010 2 2
## [84,] -0.009184101 0.2242269 1 1
## [85,] -0.068652305 0.8068993 2 2
## [86,] 0.180468131 0.1091016 1 1
## [87,] -0.165890279 0.7920560 2 2
## [88,] 0.150439388 0.1914129 1 1
## [89,] 1.283597810 0.6141207 2 2
## [90,] 0.034226004 0.2416539 1 1
## [91,] 0.938828130 0.6418170 2 2
## [92,] 0.033280984 0.2170993 1 1
## [93,] -0.259998069 0.8190769 2 2
## [94,] 0.113969217 0.1819611 1 1
## [95,] 0.840929544 0.6610181 2 2
## [96,] 0.127975410 0.2147438 1 1
## [97,] 0.451227606 0.7057840 2 2
## [98,] 0.145929302 0.2292715 1 1
## [99,] 0.035067036 0.1675377 1 1
## [100,] 0.296801022 0.7375518 2 2
sum(hidden_states==Fstates)
## [1] 100