-
Notifications
You must be signed in to change notification settings - Fork 0
/
AVAP_10_xtra_yr_interact.R
352 lines (297 loc) · 9.96 KB
/
AVAP_10_xtra_yr_interact.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
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
###set up a dataframe for the observations
## some changes
# once again trying git
setwd('~/Documents/research/alpine air photo veg change')
veg <- read.csv('Alpvegcov.csv') # read in from data folder
veg <- veg[with(veg,order(veg$Site, veg$date)), ]
veg[,3:5] <- round(veg[,3:5])
veg[,6] <- 1600 - veg[,3] - veg[,4] - veg[,5]
veg[,2] <- floor((veg$date-1926)/10)
################this recodes dates into 10yr intervals###################
names(veg[,6]) <- "other.hits"
sites <- read.csv('Alpsites.csv') # read in from data folder
#yrs <- veg$date-1970
hd <- floor((veg$date-1926)/10)
veg <- merge(veg,sites)
veg <- veg[with(veg,order(veg$Site, veg$date)), ]
veg[,7] <- floor((veg[,7]-1926)/10)
first <- min(veg[,2])
last <- max(veg[,2])
ord <- rep(1:3,times=23)
veg <- cbind(veg,ord)
veg.w <- reshape(veg, timevar="ord", idvar="Site", direction="wide")
all.yrs.a <- rep(NA,times = (23) * (last - first + 2) * 4)
dim(all.yrs.a) <- c(23,(last - first + 2),4)
#years of observations
y.o.obs <- rep(NA, times=23*3) #empty vectors
dim(y.o.obs) <- c(23,3)
y.o.obs[,1] <- veg.w[,2]#this is the year of the first observation
y.o.obs[,2] <- veg.w[,15]
y.o.obs[,3] <- veg.w[,28]
yoo1 <- y.o.obs[,1]+1
all.yrs.a <- rep(NA,times = (23) * (last - first +2) * 4)
dim(all.yrs.a) <- c(23,(last - first + 2),4)
o<- rep(NA, times=23)
for (k in 1:4) {
for (i in 1:23) {
o[i] <- veg.w[i, k+2] #this is the 1st observation for veg type k
all.yrs.a[i, y.o.obs[i,1] - first + 2, k] <- o[i]
}
for (i in 1:23) {
o[i] <- veg.w[i, k+15] #this is the 2nd observation
all.yrs.a[i, y.o.obs[i,2] - first + 2, k] <- o[i]
}
for (i in 1:23) {
o[i] <- veg.w[i, k+28] #this is the 3rd observation
all.yrs.a[i, y.o.obs[i,3] - first + 2, k] <- o[i]
}
}
# setup covariates for time since grazing, wildfires and clearing
tsg <- floor((sites$last.grazed-1926)/10)
t <- all.yrs.a[,,1]
for (i in 1:23){
for (j in 1 : (last - first + 2) ) {
t[i, j] <- ifelse( (j + first - 2) > tsg[i] , 1, 0)
}
}
# now for CB this version puts it in yr BEFORE obs.
cb.t <- t
cb.t[,1:(last - first + 2)] <- 0
for (i in 1:23) {
o[i] <- ifelse(sites[i, 3] > 0, (sites[i, 3]/100), 0) #this is the observed redcution
# cb.t[i, y.o.obs[i,1]:(y.o.obs[i,1]+1)] <- o[i] # 2 period effect of C&B
cb.t[i, y.o.obs[i,1]] <- o[i]
}
#cb1 <- cb.t
#cb.t[,2:69] <- 1
for (i in 1:23) {
o[i] <- ifelse(sites[i, 4] > 0, (sites[i, 4]/100), 0) #this is the 2nd observation
# cb.t[i, y.o.obs[i,2]:(y.o.obs[i,2]+1)] <- o[i]
cb.t[i, y.o.obs[i,2]] <- o[i]
}
# for wildfire, lets just assume 15 y before photo
fy.t <- t
fy.t[ , 1:(last - first + 2)] <- 0
for (i in 1:23) {
o[i] <- ifelse(sites[i, 5] > 0, 1, 0) #this is the 2nd observation
fy.t[i, y.o.obs[i,2] ] <- o[i]
}
#now for 2nd interval
for (i in 1:23) {
o[i] <- ifelse(sites[i, 6] > 0, 1, 0) #this is the 3rd observation
fy.t[i, y.o.obs[i,3] ] <- o[i]
}
library(R2jags)
#library(runjags)
#source('~/insects model/doJAGS.R')
KK.init <- matrix(runif(92,0,4), nrow=23)
for (i in 1:23) { KK.init[i,1] <- log(max(veg.w[i,c(3,16,29)])+2) }
for (i in 1:23) { KK.init[i,2] <- log(max(veg.w[i,c(4,17,30)])+2) }
for (i in 1:23) { KK.init[i,3] <- log(max(veg.w[i,c(5,18,31)])+2) }
for (i in 1:23) { KK.init[i,4] <- log(max(veg.w[i,c(6,19,32)])+2) }
interval <- veg.w[,15] - veg.w[,2]
interval2 <- veg.w[,28] - veg.w[,15]
dif <- veg.w[,16] - veg.w[,3]
dif <- cbind(dif, veg.w[,17] - veg.w[,4] )
dif <- cbind(dif, veg.w[,18] - veg.w[,5] )
dif <- cbind(dif, veg.w[,19] - veg.w[,6] )
dif <- cbind(dif, veg.w[,29] - veg.w[,16] )
dif <- cbind(dif, veg.w[,30] - veg.w[,17] )
dif <- cbind(dif, veg.w[,31] - veg.w[,18] )
dif <- cbind(dif, veg.w[,32] - veg.w[,19] )
N.init <- all.yrs.a
for (i in 1:23) {
for (j in 1:(last - first + 2)) {
for (k in 1:4) {
N.init[i,j,k] <- ifelse(j == y.o.obs[i,1] + 1, NA,
(ifelse(is.na(N.init[i,j,k]), max(round(veg.w[i, k+2] + dif[i,k] / interval[i] * (j - y.o.obs[i,1]+1) ),1), N.init[i,j,k] ) ) )
N.init[i,j,k] <- ifelse(j < y.o.obs[i,2] +1, N.init[i,j,k],
max(round(veg.w[i, k+15] + dif[i,k+3] / interval2[i] * (j - y.o.obs[i,2]+1)),1))
N.init[i,j,k] <- ifelse(j == y.o.obs[i,2]+1, NA, N.init[i,j,k])
N.init[i,j,k] <- ifelse(j == y.o.obs[i,3]+1, NA, N.init[i,j,k])
N.init[i,j,4] <- ifelse(is.na(N.init[i,j,4]), NA, round(N.init[i,j,4]*.7))
#for N1 NA, remove for most starts
# N.init[i,1,k] <- NA
}}}
veg.inits <- function() list(
# mn.K = c(3,3,3,3),
alpha.100 = c(5,9,0.5,1.8), #rnorm(3, mean = 0.05, sd = 0.01),
N = N.init,
# KK = KK.init, #runif(23, 500, 1500),
sd1 = runif(1,0.5,1),
sd2 = runif(1,0.5,1),
gr.ef = rnorm(4, mean = 0.5, sd = 0.1),
KT = rnorm(4, mean = 0.5, sd = 0.1),
cb.ef = rnorm(4, mean = 0.5, sd = 0.1),
f.ef = rnorm(4, mean = 0.0, sd = 0.1),
a = c(3,2.5,3.5, 3),
b = c(0,1,-1, 0)
) #THIS IS THE INITIALS LIST
#define data
veg.data <- list(
Nsites = 23,
# alt = sites[,7],
precip = sites[,8],
Nveg = 4,
T = last - first+1,
yoo1 = yoo1,
N = all.yrs.a,
sg = as.matrix(t[ , 1:(last - first + 2)]),
cb.t = as.matrix(cb.t[ , 1:(last - first + 2)]),
fy.t = as.matrix(fy.t[ , 1:(last - first + 2)])) ## ,
# THIS IS THE DATA LIST
###lets try a Ricker function
cat(' model{
for (j in 1:Nsites) {
for (t in yoo1[j] : T) {
for (k in 1:Nveg) {
N[j, t + 1, k] ~ dpois(lambda[j, t, k])
lambda[j, t, k] <- (N[j, t, k] + 0.5) * ( Er[j, t, k] ) * exp(cb.ef[k] * cb.t[j,t] + f.ef[k] * fy.t[j, t])
Er[j, t, k] <- exp( alpha.100[k] / 100 * ( 1 - N[j, t, k] / K[j, t, k]) )
log(K[j, t, k]) <- log.K[j, t, k]
log.K[j, t, k] ~ dnorm(mu.K[j,t,k], sd1^-2)#T(0,7.3) # spatio-temporal variation around the linear predicted K (grazing, fire)
mu.K[j,t,k] <- KK[j, k] + KT[k] * (t-4) + gr.ef[k] * sg[j, t] #+ f.ef[k] * fy.t[j, t]
+ tg.int[k] * (t-4) * sg[j, t]
} #close loop round veg
} #close loop round time
for (k in 1:Nveg) {
N[j,1,k] ~ dpois(start[j, k])
log(start[j, k]) <- log.start[j, k]
log.start[j, k] ~ dnorm(KK[j, k]+ KT[k] * (-3), 0.5)#T(0,7.38) #~ dnorm(4.5,.05)T(0,7.38) # + KT[k] * (1-4) #
} #close loop around veg
} #close loop around sites
for (k in 1:Nveg) {
alpha.100[k] ~ dnorm(0.05,0.7)T(0, )
gr.ef[k] ~ dnorm(0, 0.05) # for grazing removal
cb.ef[k] ~ dnorm(0, 0.05) # T(0,10)
f.ef[k] ~ dnorm(0, 0.05)
a[k] ~ dnorm(4.5,.7)#T(0,7.38)
b[k] ~ dnorm(0,0.01)
KT[k] ~ dnorm(0,0.01) #SHOULD these sum to 0?
tg.int[k] ~ dnorm(0,0.01)
c.f.int[k] ~ dnorm(0,0.01)
} # close loop around veg
for (j in 1: Nsites) {
for (k in 1: 4) {
KK[j,k] ~ dnorm(mn.K[j,k], sd2^-2)#T(0,7.38) #site level departures from predicted K from Precip
mn.K[j,k] <- a[k] + b[k] * (precip[j] - 1465)/216 # (alt[j]-1433)/185 #standardize
} # close loop round veg
} # close loop round sites
sd1 ~ dt(0, .016, 1)T(0, 7)
sd2 ~ dt(0, .016, 1)T(0, 7)
} ' , file=(modelfile <- tempfile()))
###fit model
AlpVeg10 <- jags(veg.data, veg.inits,
c( 'tg.int', 'KT','sd1','sd2', 'alpha.100', 'gr.ef', 'cb.ef', 'f.ef', 'KK', 'a', 'b'),
modelfile, n.chains=3, n.iter=40000, n.burnin=10000)
AlpVeg10.upd <- update(AlpVeg10,
params= c('tg.int','KT','sd1','sd2', 'alpha', 'gr.ef', 'cb.ef', 'f.ef', 'a', 'b' ),
n.iter=15000, n.thin=15)
print(AlpVeg10)
par(mfrow = c(4, 2))
traceplot(AlpVeg10, mfrow=c(4, 2), varname = c('tg.int', 'K', 'sd', 'gr.ef', 'cb.ef', 'f.ef', 'a', 'b'), ask=FALSE)
attach.jags(AlpVeg10.upd)
par(mfrow = c(4,2))
plot(density(a[,1]))
plot(density(a[,2]))
plot(density(a[,3]))
plot(KK[,19,1], KK[,19,2])
plot(KK[,19,3], KK[,19,2])
plot(KK[,19,1], KK[,19,3])
plot(KK[,19,3], KK[,19,4])
plot(KT[,1], gr.ef[,1])
abline(h=0,v=0)
plot(KT[,2], gr.ef[,2])
abline(h=0,v=0)
plot(KT[,3], gr.ef[,3])
abline(h=0,v=0)
plot(KT[,4], gr.ef[,4])
abline(h=0,v=0)
plot(KT[,1], tg.int[,1])
abline(h=0,v=0)
plot(KT[,2], tg.int[,2])
abline(h=0,v=0)
plot(KT[,3], tg.int[,3])
abline(h=0,v=0)
plot(KT[,4], tg.int[,4])
abline(h=0,v=0)
plot(gr.ef[,1], tg.int[,1])
abline(h=0,v=0)
plot(gr.ef[,2], tg.int[,2])
abline(h=0,v=0)
plot(gr.ef[,3], tg.int[,3])
abline(h=0,v=0)
plot(gr.ef[,4], tg.int[,4])
abline(h=0,v=0)
## cb vs b
plot(cb.ef[,1], b[,1])
abline(h=0,v=0)
plot(cb.ef[,2], b[,2])
abline(h=0,v=0)
plot(cb.ef[,3], b[,3])
abline(h=0,v=0)
plot(cb.ef[,4], b[,4])
abline(h=0,v=0)
#indep of fire and cb
plot(cb.ef[,1], f.ef[,1])
abline(h=0,v=0)
plot(cb.ef[,2], f.ef[,2])
abline(h=0,v=0)
plot(cb.ef[,3], f.ef[,3])
abline(h=0,v=0)
plot(cb.ef[,4], f.ef[,4])
abline(h=0,v=0)
plot(a[,1], b[,1])
plot(a[,1], alpha.100[,1])
plot(a[,1], cb.ef[,1])
plot(a[,1], gr.ef[,1])
plot(a[,1], f.ef[,1])
plot(cb.ef[,1], f.ef[,1])
plot(gr.ef[,1], f.ef[,1])
plot(alpha.100[,1], b[,1])
plot(alpha.100[,1], cb.ef[,1])
plot(alpha.100[,1], gr.ef[,1])
plot(alpha.100[,1], f.ef[,1])
plot(a[,1], KK[,19,1])
plot(b[,1], KK[,19,1])
plot(alpha.100[,1], KK[,19,1])
plot(cb.ef[,1], KK[,19,1])
plot(gr.ef[,1], KK[,19,1])
plot(f.ef[,1], KK[,19,1])
plot(a[,2], b[,2])
plot(a[,2], alpha.100[,2])
plot(a[,2], cb.ef[,2])
plot(a[,2], gr.ef[,2])
plot(a[,2], f.ef[,2])
plot(cb.ef[,2], f.ef[,2])
plot(gr.ef[,2], f.ef[,2])
plot(alpha.100[,2], b[,2])
plot(alpha.100[,2], cb.ef[,2])
plot(alpha.100[,2], gr.ef[,2])
plot(alpha.100[,2], f.ef[,2])
plot(a[,2], KK[,19,2])
plot(b[,2], KK[,19,2])
plot(alpha.100[,2], KK[,19,2])
plot(cb.ef[,2], KK[,19,2])
plot(gr.ef[,2], KK[,19,2])
plot(f.ef[,2], KK[,19,2])
plot(a[,3], b[,3])
plot(a[,3], alpha.100[,3])
plot(a[,3], cb.ef[,3])
plot(a[,3], gr.ef[,3])
plot(a[,3], f.ef[,3])
plot(cb.ef[,3], f.ef[,3])
plot(gr.ef[,3], f.ef[,3])
plot(alpha.100[,3], cb.ef[,3])
plot(alpha.100[,3], gr.ef[,3])
plot(alpha.100[,3], f.ef[,3])
plot(a[,3], KK[,19,3])
plot(b[,3], KK[,19,3])
plot(alpha.100[,3], KK[,19,3])
plot(cb.ef[,3], KK[,19,3])
plot(gr.ef[,3], KK[,19,3])
plot(f.ef[,3], KK[,19,3])
plot(density(sd[, 1]))
plot(density(alpha.100[, 1]))
plot(density(gr.ef[, 1]))
plot(density(cb.ef[, 1]))