-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsimulation.R
186 lines (147 loc) · 6.21 KB
/
simulation.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
simulation <- function(input_file, output_dir, tr, ir, hr) {
tr <<- tr
cat("Simulating synthetic image...\n")
cell_data <- read.csv(input_file)
input_file_name <- basename(tools::file_path_sans_ext(input_file))
filename <- file.path(output_dir, paste0(input_file_name, "_model_TR_", tr, "_IR_", ir, "_HR_", hr, ".Rda"))
load(file = filename)
W <<- owin(c(min(cell_data[[1]]), max(cell_data[[1]])),
c(min(cell_data[[2]]), max(cell_data[[2]])))
cell_type_list <<- sort(unique(cell_data[[3]]))
cell_type_num <<- length(cell_type_list)
G <- list()
coef_names <- names(coef)
for (i in seq_len(cell_type_num)) {
coef_current <- c()
for (r in seq(ir, tr, ir)) {
for (j in seq_len(cell_type_num)) {
mark1 <- cell_type_list[i]
mark2 <- cell_type_list[j]
interaction_name <- if (i <= j) {
paste('X', mark1, 'x', 'X', mark2, 'x', r, sep = '')
} else {
paste('X', mark2, 'x', 'X', mark1, 'x', r, sep = '')
}
coef_current <- c(coef_current, coef[grep(interaction_name, coef_names, value = FALSE)])
}
}
G[[i]] <- coef_current
}
G <<- G
B <- c()
for (i in seq_len(cell_type_num)) {
B_current <- if (i == 1) {
exp(coef[i])
} else {
exp(coef[i] + coef[1])
}
B <- c(B, B_current)
}
B <<- B
resample_percent <- 100
model_mark_prob <- (dplyr::count(cell_data, marks)$n) / nrow(cell_data)
intensity <- nrow(cell_data) / area(W)
cell_pattern_random <- rpoispp(intensity, win = W)
pattern_exist <- data.frame(cell_pattern_random)
pattern_exist[, 3] <- as.factor(sample(cell_type_list, nrow(pattern_exist), replace = TRUE, model_mark_prob))
cell_pattern_num <- nrow(pattern_exist)
resample_num <- ceiling(cell_pattern_num * resample_percent / 100)
if (resample_num != 0) {
for (i in seq_len(resample_num)) {
point_current_idx <- sample(1:cell_pattern_num, 1)
point_current <- pattern_exist[point_current_idx, ]
counts_current <- t(apply(point_current[,1:2], 1, calc_dist, pattern_exist))
mark_prob_current <- apply(counts_current, 1, calc_mark_prob)
point_current[, 3] <- mark_prob_current
pattern_exist[point_current_idx, ] <- point_current
}
}
marked_simulated_pattern <- ppp(c(pattern_exist[,1]), c(pattern_exist[,2]), window = W, marks = factor(pattern_exist[,3], levels = cell_type_list))
# filename <- file.path(output_dir, paste0(input_file_name, "_synthetic_pattern_TR_", tr, "_IR_", ir, "_HR_", hr, ".Rda"))
# save(marked_simulated_pattern, file = filename)
synthetic_image_vis(data.frame(marked_simulated_pattern), input_file_name, output_dir, tr, ir, hr)
cat("Simulation completed and synthetic image saved!\n")
}
truncate_polygon <- function(polygon, center, cutoff_distance) {
set.seed(3)
# Convert data to spatial objects
pol <- SpatialPolygons(list(Polygons(list(Polygon(polygon)), ID = 1)))
center <- t(as.matrix(center))
pt <- SpatialPoints(center)
buffered_point <- gBuffer(pt, width = cutoff_distance)
intersection <- gIntersection(pol, buffered_point)
if (!is.null(intersection)) {
return(as(intersection, "SpatialPolygons"))
} else {
return(NULL)
}
}
synthetic_image_vis <- function(pattern_df, input_file_name, output_dir, tr, ir, hr){
colors <- rainbow(cell_type_num)
plot_base <- ggplot() +
theme_void() +
theme(legend.position = "none", plot.background = element_rect(fill = "black"))+ scale_fill_identity()
buffer_distance <- 20
points <- as.matrix(pattern_df[, 1:2])
marks <- as.integer(pattern_df$mark)
vor <- deldir(points)
tile_list <- tile.list(vor)
plot_data_list <- list()
suppressWarnings({
for (i in 1:length(tile_list)) {
tile <- tile_list[[i]]
polygon <- tile[c("x", "y")]
truncated <- truncate_polygon(polygon, points[i,], buffer_distance)
if (!is.null(truncated)) {
temp_data <- fortify(truncated)[,c("long", "lat")]
temp_data$color <- as.character(colors[marks[i]+1])
temp_data$group <- i
plot_data_list[[i]] <- temp_data
}
}
})
# Combine all the data frames into one
plot_data <- do.call(rbind, plot_data_list)
# Plot all polygons at once
plot_base <- plot_base +
geom_polygon(data = plot_data, aes(x = long, y = lat, fill = color, group = group), color = "white", size = 0.3)
filename <- file.path(output_dir, paste0(input_file_name, "_synthetic_image_TR_", tr, "_IR_", ir, "_HR_", hr, ".png"))
ggsave(filename = filename, dpi = 500, bg = "black", width=10, height=10)
}
get_random_sampled_marks <- function(types, num, prob) {
sampled_marks <- sample(types, num, TRUE, prob)
return(sampled_marks)
}
calc_dist <- function(new_points, existing_points) {
new_points <- data.frame(t(new_points))
x <- new_points[1, 1]
y <- new_points[1, 2]
xmin <- max(0, x-tr)
ymin <- max(0, y-tr)
xmax <- min(W$xrange[2], x+tr)
ymax <- min(W$yrange[2], y+tr)
existing_points_within_range <- existing_points[which(existing_points[,1] > xmin & existing_points[,1] < xmax & existing_points[,2] > ymin & existing_points[,2] < ymax),]
eu_dist <- dist(new_points, existing_points_within_range[,1:2])
existing_points_within_range <- cbind(existing_points_within_range, c(eu_dist))
existing_points_within_range <- existing_points_within_range[which(existing_points_within_range[,4] < 500 & existing_points_within_range[,4] > 0), 3:4]
mark_count <- matrix(0, nrow = 5, ncol = 5)
for (i in seq_len(nrow(existing_points_within_range))) {
mark_count_row <- as.integer(existing_points_within_range[i, 1])
mark_count_col <- existing_points_within_range[i, 2] %/% 100 + 1
mark_count[mark_count_row, mark_count_col] <- mark_count[mark_count_row, mark_count_col] + 1
}
return(mark_count)
}
calc_mark_prob <- function(counts) {
m_prob <- c()
for (m in cell_type_list) {
m_int <- as.integer(m) + 1
G_type <- G[[m_int]]
counts_type <- counts
interact_term <- exp(sum(G_type * counts_type))
intensity_term <- B[m_int]
m_prob <- c(m_prob, unname(intensity_term * interact_term))
}
m_prob <- m_prob / sum(m_prob)
return(factor(sample(cell_type_list, 1, prob = m_prob), levels = cell_type_list))
}