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
|
debugOutput <- FALSE
# Slash direction for directory creation
platform = "linux"
# Possibly Spp names info
# Required Packages
require(rpart)
require(mgcv)
require(gbm)
require(fields)
require(splancs)
require(maps)
require(maptools)
# Source STEM Libraries
source(paste(code.directory,"stem.library.A.R",sep=""))
source(paste(code.directory,"stem.library.B.R",sep=""))
source(paste(code.directory,"stem.library.ST.R",sep=""))
# Source MORE STEM Functions
#source(paste(code.directory,"STEM_erd.data.creation.R",sep=""))
source(paste(code.directory,"STEM_erd.srd.data.subsetting.R",sep=""))
source(paste(code.directory,"STEM_ModelFitting.R",sep="") )
source(paste(code.directory,"predict.eBird.ST.matrix.R",sep=""))
source(paste(code.directory,"STEM_st.matrix.cv.average.R",sep=""))
## Name of space time matrix sub-directory
st.matrix.name <- "stm"
## Name of space time matrix ave map sub-directory
st.matrix.ave.maps <- "stm.ave.maps"
# ------------------------
# Select One Species
# speciesList - a list of vectors. vector[1] = scientificName & vector[2] = commonName
# ------------------------
speciesList <- list(
c("Hylocichla_mustelina", "Wood_Thrush"))
#c("Cyanocitta_cristata", "Blue_Jay"))
#c("Euphagus_carolinus", "Rusty_Blackbird"))
#c("Turdus_migratorius", "American_Robin"))
#c("Seiurus_aurocapilla", "Ovenbird"),
#c("Vireo_olivaceus", "Red-eyed_Vireo"))
#c("Cardinalis_cardinalis", "Northern_Cardinal"))
#c("Junco_hyemalis", "Dark-eyed_Junco"),
#c("Piranga_olivacea", "Scarlet_Tanager"))
#c("Carduelis_pinus", "Pine_Siskin"))
#c("Hylocichla_mustelina", "Wood_Thrush"))
#c("Colinus_virginianus", "Northern_Bobwhite"))
# -----------------------------------------------------
# SRD maximum sample size - this further restricts SRD
# in addition to ERD spatial restrictions above)
# -----------------------------------------------------
srd.max.sample.size <- 5000
## response.family = "bernoulli" or "gaussian"
response.family <- "bernoulli"
# ----------------------------------------------------------------------
# Spatial Extent by Point in Rectangle, Polygon, or ShapeFile
# ----------------------------------------------------------------------
#spatial.extent.list <- NULL # Full US
# -----------
#spatial.extent.list <- list(type = "rectangle",
# lat.min = 25.0,
# lat.max = 50.0,
# lon.min = -130.0,
# lon.max = -65.0)
# -----------
spatial.extent.list <- list(type = "polygon",
# data.frame with named vertices x and y
polygon.vertices =
data.frame(x = c(-85,-85,-70,-70),
y = c(35, 50, 50, 35)) )
# -----------
#spatial.extent.list <- list(type="shapefile",
# shape.dir = data.dir.shapefiles,
# att.selection.column.name = "BCR",
# # watch capitalization!
# shape.filename = "bcr.shp",
# selected.shape.names = c(13))
# -----------
#spatial.extent.list <- list(type="shapefile",
# shape.dir = data.dir.shapefiles,
# att.selection.column.name = "STATE_NAME",
# shape.filename = "STATES.shp",
# selected.shape.names = c("New York"))
# -------------------------------------------------------------------
# Temporal Extent
# -------------------------------------------------------------------
temporal.extent.list <- list(begin.jdate = 1,
begin.year = 2006,
end.jdate = 366,
end.year = 2008)
iii.year <- 2008
n.intervals.per.year <-52
n.intervals <- n.intervals.per.year + 1
jdate.seq <- seq(from = temporal.extent.list[['begin.jdate']], to = temporal.extent.list[['end.jdate']], length = (n.intervals))
if (n.intervals > 1) {
jdate.seq <- round((jdate.seq[2:n.intervals] + jdate.seq[1:(n.intervals-1)])/2)
}
year.seq <- rep(iii.year, length(jdate.seq))
begin.seq <- 1
end.seq <- length(jdate.seq)
# ------------------------
# Select predictors - must be a subset of what is in erd…..
# ------------------------
predictor.names <- c("YEAR", "TIME", "DAY",
"EFFORT_HRS", "EFFORT_DISTANCE_KM",
"NUMBER_OBSERVERS", "HOUSING_DENSITY",
"ELEV_NED",
"CAUS_TEMP_AVG",
"CAUS_TEMP_MIN", "CAUS_TEMP_MAX",
"CAUS_PREC", "CAUS_SNOW",
"NLCD2001_FS_C11_750_PLAND", "NLCD2001_FS_C12_750_PLAND",
"NLCD2001_FS_C21_750_PLAND", "NLCD2001_FS_C22_750_PLAND",
"NLCD2001_FS_C23_750_PLAND", "NLCD2001_FS_C24_750_PLAND",
"NLCD2001_FS_C31_750_PLAND", "NLCD2001_FS_C41_750_PLAND",
"NLCD2001_FS_C42_750_PLAND", "NLCD2001_FS_C43_750_PLAND",
"NLCD2001_FS_C52_750_PLAND", "NLCD2001_FS_C71_750_PLAND",
"NLCD2001_FS_C81_750_PLAND", "NLCD2001_FS_C82_750_PLAND",
"NLCD2001_FS_C90_750_PLAND", "NLCD2001_FS_C95_750_PLAND")
# -------------------------------------------------------------------
# Data Splitting Parameters
# -------------------------------------------------------------------
split.par.list <- list(grid.cell.min.lat = 0.75,
grid.cell.min.lon = 0.75,
min.val.cell.locs = 2,
fraction.training.data=0.8,
mfrac=1.0,
plot.it=FALSE)
# -------------------------------------------------------------------
# CV Fold Splitting Parameters
# -------------------------------------------------------------------
cv.folds <- 1
cv.list <- c(1:cv.folds)
cv.folds.par.list <- list(grid.cell.min.lat = 0.75,
grid.cell.min.lon = 0.75,
min.val.cell.locs = 2,
fraction.training.data = 0.8,
mfrac = 1.0,
plot.it = FALSE)
# ---------------------------------------------------------------------
# STEM Ensemble Design
# ---------------------------------------------------------------------
iii.scale <- 2 # largest scale in STEM MS ==> current magic value
stem.init.par.list <- list(
# --------------------------------------------------
# Spatial Design
# --------------------------------------------------
spatial.extent.list = NULL,
spatial.region.par.list = list(
# Number of MC regionalizations for each time interval
n.mc.regions=1,
# define minimum area & sample size of rectangles
regional.cell.min.lat = 3.00*iii.scale,
regional.cell.min.lon = 4.00*iii.scale,
# Underlying Coverage Grid
# Older declarations
# grid.cell.min.lat = 1*(iii.scale^0.5), #5 degrees
# grid.cell.min.lon = 1.75*(iii.scale^0.5), # 10 degrees
# Simplified
grid.cell.min.lat = 3.00*iii.scale/3, #5 degrees
grid.cell.min.lon = 4.00*iii.scale/3, # 10 degrees
n.centers.per.region = 1, # Density Dependent!!!
min.data.size = 10),
# --------------------------------------------------
# Temporal Design:
# Slice year into n.intervals prediction points
# --------------------------------------------------
n.intervals=60, #add 1! # 81 for migration
begin.jdate=1 ,
end.jdate = 366,
season.window.width=40, #40 # Fitting window
prediction.window.width=36, #36
# --------------------------------------------------
# Sampling
# --------------------------------------------------
sampling.par.list = list(
split.by.location = FALSE,
# if TRUE uses locs == ensemble.data.locs
ST.basis.rotation = FALSE,
# if TRUE assumes {lat,lon} ={$y, $ x}
# and julian date == $JDATES
p.train=0.75))
base.model.par.list <- rpart.control(cp = 0.01, xval = 0, minbucket = 10)
# ---------------------------------------------------------------------
# conditioning variables
# ---------------------------------------------------------------------
conditioning.vars <- list(EFFORT_HRS = 1.0,
EFFORT_DISTANCE_KM = 1.0,
TIME = 7.0,
NUMBER_OBSERVERS = 1.0,
I.STATIONARY = 0)
# -----------------------------------------------------------------------------
# -----------------------------------------------------------------------------
# ST Matrix Maps
# (AKA Halloween Maps)
# -----------------------------------------------------------------------------
# -----------------------------------------------------------------------------
ns.rows <- 40
map.plot.pixel.width <- 1000
z.max <- 0.50
z.min <- 0.0
map.filename.tag <- "demo."
state.map <- TRUE
state.map.lwd <- 1.0
county.map <- TRUE
county.map.lwd <- 0.5
pred.grid.size <- ns.rows
## map.plot.width <- map.plot.pixel.width
map.plot.width <- 1000
## This is duplicated in stem.init.par.list
spatial.extent.list <- NULL
map.tag <- map.filename.tag
date.bar <- TRUE
print.date <- TRUE
## KFW resp.transformation.code <- NULL
|
