-
Notifications
You must be signed in to change notification settings - Fork 1
/
RunN15InverseRW.R
240 lines (203 loc) · 6.91 KB
/
RunN15InverseRW.R
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
78
79
80
81
82
83
84
85
86
87
88
89
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
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
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
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
#Note that first you have to set the directory: Session > Set Working Directory
rm(list=ls()) # Housekeeping
## Load libraries
library(R.matlab) # Used to read in/out matlab binary files
library(openxlsx) # Used to read xlsx files
library(limSolve) # The original LIM Package
library(MASS) # Matrix library
source('xsampleN15.r') # The N15 xsample algorithm
source('ExternalFunctions.r') # The place for ancillary functions
###############################
#### 1. Read in input data ####
###############################
Model <- 'NEMURO.Offshore'
datafile <- paste('N15InverseModelRW.',Model,'.RW.mat',sep="")
data <- readMat(datafile, fixNames=TRUE)
h <- data[['h']]
b <- data[['b']]
ba <- data[['ba']]
be <- data[['be']]
Ae <- data[['Ae']]
Aa <- data[['Aa']]
G <- data[['G']]
sdba <- data[['sdba']]
InputCol <- data[['InputCol']]
A <- data[['A']]
Inputs <- data[['Inputs']]
###############################
#### 2. Set run parameters ####
###############################
KeeperRatio <- 10000
IterLength <- 10000000
OutputLength <- IterLength / KeeperRatio
BurninLength <- IterLength / 100
## Set based on which forward model
if (Model == "DIAZO.Coastal"){
jmpLength <- 0.02
}else if (Model == "DIAZO.Mesohaline") {
jmpLength <- 0.005
}else if (Model == "NEMURO.Coastal") {
jmpLength <- 0.001
}else if (Model == "NEMURO.Offshore") {
jmpLength <- 0.0001
}
jmpLength15 <- 0.02
## Cut to isolate non-15N equations (11 -> 15 here)
Aa2 <- Aa[11:14,]
ba2 <- ba[11:14]
sdba2 <- sdba[11:14]
d15N0 <- c(0, 0, 0, 0, 0, 0)
## Read N15 values from forward model
upNO315N <- Inputs[8, InputCol]
Mes15N <- Inputs[9, InputCol]
Det15N <- Inputs[7, InputCol]
Doc15N <- Inputs[22, InputCol]
del15Nknown <- c(upNO315N, Mes15N, Doc15N, Det15N)
## L2MN Solution (without N15)
# Start
test <- lsei(A = Aa2, B = ba2, E = Ae, F = be, G = G, H = h, type=2) # No N15 here
X <- test[['X']]
solNorm <- test[['solutionNorm']]
lseisol <- as.matrix(X)
rm(test, X) # Housekeeping
#Central Value Solution
center <- xranges(E = Ae, F = be, G = G, H = h,
ispos = FALSE, tol = 1e-8, central = TRUE, full=FALSE)
centralval <- center[,3]
##############################################################
#### 3a. Iterative burn-in in order to establish matricies ####
##############################################################
## First burnin
test2 <- xsample(A = Aa2, B = ba2, E = Ae, F = be, G = G,
H = h, sdB = sdba2*10, iter = BurninLength/2,
type="mirror", jmp=runif(1, 0, 1)*jmpLength/5,
x0 = centralval, fulloutput='TRUE')
Burninmat <- test2[['X']]
Startpt <- Burninmat[BurninLength/2,]
rm(Burninmat) # Housekeeping
## Second burnin
test2 <- xsample(A = Aa2, B = ba2, E = Ae, F = be, G = G,
H = h, sdB = sdba2, iter = BurninLength/2, type="mirror",
jmp=jmpLength+runif(1, 0, 1)*jmpLength/5, x0 = Startpt,
fulloutput='TRUE')
Burninmat <- test2[['X']]
Startpt <- Burninmat[BurninLength/2,]
rm(Burninmat) # Housekeeping
############################################
#### 3b. Run the full model without N15 ####
############################################
test2 <- xsample(A = Aa2, B = ba2, E = Ae, F = be, G = G,
H = h, sdB = sdba2, iter = IterLength, outputlength = OutputLength,
type="mirror", jmp=jmpLength+runif(1, 0, 1)*jmpLength/5, x0 = Startpt,
fulloutput='TRUE')
MCMCmatplain <- test2[['X']]
##############################
#### 4. N15 model section ####
##############################
# Use full model solution to initialize the N15 matricies
MCMCmatmean <- colMeans(MCMCmatplain)
RN2 <- 0.0036765;
Eps_Remin <- -1
Eps_Eg <- -2
Eps_NH4up <- -10
Alpha_NH4up <- exp(Eps_NH4up/1000)
Alpha_Eg <- exp(Eps_Eg/1000)
Alpha_Remin <- exp(Eps_Remin/1000)
sdbafactor <- (RN2*Alpha_Remin - RN2)/10
sdba[1] <- sum(MCMCmatmean[c(1,3,10)]) * sdbafactor
sdba[2] <- sum(MCMCmatmean[c(4,11,18,21,25,32)]) * sdbafactor
sdba[3] <- sum(MCMCmatmean[c(2,3,4,5,6,7,8)]) * sdbafactor
sdba[4] <- sum(MCMCmatmean[c(9,10,11,12,13,14,15)]) * sdbafactor
sdba[5] <- sum(MCMCmatmean[c(12,16,17,18,19,20,28)]) * sdbafactor
sdba[6] <- sum(MCMCmatmean[c(5,13,16,21,22,23,24,29)]) * sdbafactor
sdba[7] <- sum(MCMCmatmean[c(6,17,22,25,26,27,30,34)]) * sdbafactor
sdba[8] <- sum(MCMCmatmean[c(7,14,19,23,26,28,29,30,31,33)]) * sdbafactor
sdba[9] <- sum(MCMCmatmean[c(8,15,20,24,27,31,32)]) * sdbafactor
sdba[10] <- sum(MCMCmatmean[c(1,2,9,33,34)]) * sdbafactor
## Burnin - N15 (iterative process)
test2 <- xsampleN15(A = Aa, B = ba, E = Ae, F = be, G = G,
H = h, sdB = sdba*10, iter = BurninLength/2,
type="mirror", jmp=runif(1, 0, 1)*jmpLength/5,
jmp2=jmpLength15/5, x0 = centralval, del15N1=d15N0,
del15Nknown=del15Nknown, fulloutput='TRUE')
Burninmat <- test2[['X']]
del15Ntemp <- test2[['del15Ntrack']]
d15N0 <- del15Ntemp[BurninLength/2,]
Startpt <- Burninmat[BurninLength/2,]
rm(del15Ntemp, Burninmat) # Housekeeping
## Second Burnin - N15
test2 <- xsampleN15(A = Aa, B = ba, E = Ae, F = be, G = G,
H = h, sdB = sdba, iter = BurninLength/2,
type="mirror", jmp=runif(1, 0, 1)*jmpLength/5,
jmp2=jmpLength15/5, x0 = Startpt, del15N1=d15N0,
del15Nknown=del15Nknown, fulloutput='TRUE')
Burninmat <- test2[['X']]
del15Ntemp <- test2[['del15Ntrack']]
d15N0 <- del15Ntemp[BurninLength/2,]
Startpt <- Burninmat[BurninLength/2,]
rm(del15Ntemp, Burninmat)
################################
#### 5. Full model Run #########
################################
test2 <- xsampleN15(A = Aa, B = ba, E = Ae, F = be, G = G,
H = h, sdB = sdba, iter = IterLength, outputlength = OutputLength,
type="mirror", jmp=jmpLength+runif(1, 0, 1)*jmpLength/5,
jmp2=jmpLength15, x0 = Startpt, del15N1=d15N0, del15Nknown=del15Nknown,
fulloutput='TRUE')
MCMCmatN15 <- test2[['X']]
del15N <- test2[['del15Ntrack']]
randomnumber <- test2[['randomnumber']]
acceptedratio <- test2[['acceptedratio']]
##############################
#### 6. Save the output ######
##############################
fileout.mat <- paste('N15InverseModel.',Model,'.ROutputs.mat',sep="")
fileout.r <- paste('N15InverseModel.',Model,'.ROutputs.rdata',sep="")
save(file=fileout.r,
list(MCMCmatN15 = MCMCmatN15,
MCMCmatplain = MCMCmatplain,
del15N = del15N,
Aa = Aa,
ba = ba,
Ae = Ae,
be = be,
G = G,
h = h,
sdba = sdba,
lseisol = lseisol,
Inputs = Inputs,
InputCol = InputCol,
jmpLength = jmpLength,
jmpLength15 = jmpLength15,
acceptedratio = acceptedratio,
Startpt = Startpt,
centralval = centralval,
KeeperRatio = KeeperRatio,
Aa2 = Aa2,
ba2 = ba2,
sdba2 = sdba2)
)
writeMat(fileout.mat,
MCMCmatN15 = MCMCmatN15,
MCMCmatplain = MCMCmatplain,
del15N = del15N,
Aa = Aa,
ba = ba,
Ae = Ae,
be = be,
G = G,
h = h,
sdba = sdba,
lseisol = lseisol,
Inputs = Inputs,
InputCol = InputCol,
jmpLength = jmpLength,
jmpLength15 = jmpLength15,
acceptedratio = acceptedratio,
Startpt = Startpt,
centralval = centralval,
KeeperRatio = KeeperRatio,
Aa2 = Aa2,
ba2 = ba2,
sdba2 = sdba2
)