-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathDataSimulation.R
More file actions
133 lines (98 loc) · 5.71 KB
/
Copy pathDataSimulation.R
File metadata and controls
133 lines (98 loc) · 5.71 KB
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
library(simstudy)
set.seed(123) #for reproducibility
#Creating simulated covariates
site <- defData( varname = "Age", dist = "normal", formula = 75 , variance = 25 , id = "id")
site <- defData(site, varname = "Sex", dist = "categorical", formula = ".5 ;.5", id = "id")
site <- defData(site, varname = "Education", dist = "normal", formula = 16, variance = 3, id = "id")
site <- defData(site, varname = "ApoE", dist = "categorical", formula = ".5;.5", id = "id")
site <- genData(300, site)
#Adding random effects
site <- addCorData(site, idname = "id", mu = c(0), sigma = c(10), rho = .5, corstr = "cs", cnames = c("a0"))
site <- addPeriods(site, nPeriods = 4, idvars = "id", timevarName = "Age")
site$Age <- site$Age + site$period
site1 <- site2 <- site3 <- site
#Create simulated outcome based on covariates
# Site1
ROI1.sitea <- defDataAdd(varname = "ROI1",
formula = "(10000 + a0) + -250*Sex + 20*Education + 350*ApoE + (-.85)*(Age^2)",
variance = 600000)
ROI2.sitea <- defDataAdd(varname = "ROI2",
formula = "(1700 + a0) + -29.45*Sex + 6*Education + 15*ApoE + (-.095)*(Age^2)",
variance = 18000)
ROI3.sitea <- defDataAdd(varname = "ROI3",
formula = "(50000 + a0) + -350.45*Sex + 45*Education + 150*ApoE + (-2.95)*(Age^2)",
variance = 6000000)
ROI4.sitea <- defDataAdd(varname = "ROI4",
formula = "(20000 + a0) + -150.45*Sex + 15*Education + 75*ApoE + (-1.15)*(Age^2)",
variance = 1000000)
site1 <- addColumns(ROI1.sitea, site1)
site1 <- addColumns(ROI2.sitea, site1)
site1 <- addColumns(ROI3.sitea, site1)
site1 <- addColumns(ROI4.sitea, site1)
# Site2
ROI1.siteb <- defDataAdd(varname = "ROI1",
formula = "(10000 + a0) + -250*Sex + 20*Education + 350*ApoE + (-.85)*(Age^2) + 450",
variance = 750000)
ROI2.siteb <- defDataAdd(varname = "ROI2",
formula = "(1700 + a0) + -29.45*Sex + 6*Education + 15*ApoE + (-.095)*(Age^2) + 150 ",
variance = 12500)
ROI3.siteb <- defDataAdd(varname = "ROI3",
formula = "(50000 + a0) + -350.45*Sex + 45*Education + 150*ApoE + (-2.95)*(Age^2) + 1500",
variance = 5000000)
ROI4.siteb <- defDataAdd(varname = "ROI4",
formula = "(20000 + a0) + -150.45*Sex + 15*Education + 75*ApoE + (-1.15)*(Age^2) + 2500",
variance = 840000)
site2 <- addColumns(ROI1.siteb, site2)
site2 <- addColumns(ROI2.siteb, site2)
site2 <- addColumns(ROI3.siteb, site2)
site2 <- addColumns(ROI4.siteb, site2)
# Site3
ROI1.sitec <- defDataAdd(varname = "ROI1",
formula = "(10000 + a0) + -250*Sex + 20*Education + 350*ApoE + (-.85)*(Age^2) - 675",
variance = 30000)
ROI2.sitec <- defDataAdd(varname = "ROI2",
formula = "(1700 + a0) + -29.45*Sex + 6*Education + 15*ApoE + (-.095)*(Age^2) - 120 ",
variance = 10005)
ROI3.sitec <- defDataAdd(varname = "ROI3",
formula = "(50000 + a0) + -350.45*Sex + 45*Education + 150*ApoE + (-2.95)*(Age^2) - 1750",
variance = 7000000)
ROI4.sitec <- defDataAdd(varname = "ROI4",
formula = "(20000 + a0) + -150.45*Sex + 15*Education + 75*ApoE + (-1.15)*(Age^2) - 1500",
variance = 900000)
site3 <- addColumns(ROI1.sitec, site3)
site3 <- addColumns(ROI2.sitec, site3)
site3 <- addColumns(ROI3.sitec, site3)
site3 <- addColumns(ROI4.sitec, site3)
site2$id <- site2$id + 300
site3$id <- site3$id + 600
site1$STUDY <- "A"
site2$STUDY <- "B"
site3$STUDY <- "C"
full.simulated.data <- rbind(site1, site2, site3)
full.simulated.data <- full.simulated.data[,c("ROI1", "ROI2", "ROI3", "ROI4", "Age", "Sex", "ApoE", "Education", "STUDY", "id")]
full.simulated.data$Sex <- factor(full.simulated.data$Sex)
full.simulated.data$ApoE <- factor(full.simulated.data$ApoE)
full.simulated.data$STUDY <- factor(full.simulated.data$STUDY)
full.simulated.data.cs <- full.simulated.data[!duplicated(full.simulated.data$id),]
#Feature Data
feature.data <- full.simulated.data.cs[,c("ROI1", "ROI2", "ROI3", "ROI4")]
covariate.data <- full.simulated.data.cs[,c("Age","Sex", "ApoE", "Education", "STUDY")]
feature.data.long <- full.simulated.data[,c("ROI1", "ROI2", "ROI3", "ROI4")]
covariate.data.long <- full.simulated.data[,c("Age","Sex", "ApoE", "Education", "STUDY")]
if(FALSE) {
#Plots Preharm
ggplot(full.simulated.data.cs, aes(x=Age, y=ROI1, colour=STUDY)) + geom_point() + geom_smooth(method = "gam", formula = y ~s(x))
ggplot(full.simulated.data.cs, aes(x=Age, y=ROI2, colour=STUDY)) + geom_point() + geom_smooth(method = "gam", formula = y ~s(x))
ggplot(full.simulated.data.cs, aes(x=Age, y=ROI3, colour=STUDY)) + geom_point() + geom_smooth(method = "gam", formula = y ~s(x))
ggplot(full.simulated.data.cs, aes(x=Age, y=ROI4, colour=STUDY)) + geom_point() + geom_smooth(method = "gam", formula = y ~s(x))
#Harmonize
harmfeats <- ComGamHarm(feature.data = feature.data,
covar.data = covariate.data,
eb = TRUE,
parametric = TRUE,
smooth.terms = c("Age"),
k.val = 5)
harmonized.features <- as.data.frame(t(harmfeats$harm.results))
harm.data <- cbind(harmonized.features, covariate.data)
ggplot(harm.data, aes(x=Age, y=ROI4, colour=STUDY)) + geom_point() + geom_smooth(method = "gam", formula = y ~s(x), se=FALSE)
}