Among Us VGA.Rmd

---
title: "Among Us VGA"
output: rmarkdown::github_document
---


Clear 
```{r}
rm(list = ls())
```


```{r message=FALSE, warning=FALSE, include=FALSE}
#knitr::opts_chunk$set(include = FALSE, warning = FALSE)
#install.packages("remotes")
#Sys.setenv("R_REMOTES_NO_ERRORS_FROM_WARNINGS" = "true")
#remotes::install_github("orange-vertex/rdepthmap")
library(rdepthmap)
library(rgeos)
library(rgdal)
library(maptools)
library(devtools)
library(dplyr)
library(tidyverse)
library(sjmisc)
library(seqHMM) 
library(ggplot2)
library(spatstat)
library(reshape2)
library(networkD3)
```

# Process VGA Files

Reading MIF Line Files
```{r}
map_w_furniture = readOGR("Maps_Analysis/1_With_Furniture.mif", verbose = FALSE)
map_wo_furniture = readOGR("Maps_Analysis/2_Without_Furniture.mif", verbose = FALSE)
#map_ac = readOGR("Maps_Analysis/3_ac.dxf", verbose = FALSE)
#plot(map_w_furniture)
#plot(map_wo_furniture)
#plot(map_ac)
```
Make Grid on Depthmap and Conduct Basic VGA Processes
```{r fig.width=8}
#Map1
rdepthmap::importLines(map_w_furniture, "Maps_Analysis/1_With_Furniture.graph")
rdepthmap::createGrid("Maps_Analysis/1_With_Furniture.graph", gridSize = 700)
rdepthmap::fillGrid("Maps_Analysis/1_With_Furniture.graph", fillX = -39506, fillY = -9412)
rdepthmap::makeVGAGraph("Maps_Analysis/1_With_Furniture.graph")
gallery_w_furniture = rdepthmap::getPointmapData("Maps_Analysis/1_With_Furniture.graph")$map
#names(gallery_w_furniture)

#Map2
rdepthmap::importLines(map_wo_furniture, "Maps_Analysis/2_Without_Furniture.graph")
rdepthmap::createGrid("Maps_Analysis/2_Without_Furniture.graph", gridSize = 700)
rdepthmap::fillGrid("Maps_Analysis/2_Without_Furniture.graph", fillX = -39506, fillY = -9412)
rdepthmap::makeVGAGraph("Maps_Analysis/2_Without_Furniture.graph")
gallery_wo_furniture = rdepthmap::getPointmapData("Maps_Analysis/2_Without_Furniture.graph")$map
#names(gallery_wo_furniture)

#Map3
#rdepthmap::importLines(map_ac, "Maps_Analysis/3_Accessibility.graph")
#rdepthmap::createGrid("Maps_Analysis/3_Accessibility.graph", gridSize = 700)
#rdepthmap::fillGrid("Maps_Analysis/3_Accessibility.graph", fillX = -53601, fillY = -9458)
#rdepthmap::makeVGAGraph("Maps_Analysis/3_Accessibility.graph")
gallery_ac = rdepthmap::getPointmapData("Maps_Analysis/3_Accessibility.graph")$map
#names(gallery_ac)
```
Test Plot
```{r}
plot(gallery_w_furniture[, "Point.First.Moment"])
plot(gallery_wo_furniture[, "Point.First.Moment"])
plot(gallery_ac[, "Point.First.Moment"])
```
1) Process VGA Analysis : map with furniture
```{r}
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-local", radii = c("2,3,4,5,6,7"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("2"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("3"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("4"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("5"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("6"))
rdepthmap::VGA("Maps_Analysis/1_With_Furniture.graph", vgaMode = "visibility-global", radii = c("7"))
gallery_w_furniture = rdepthmap::getPointmapData("Maps_Analysis/1_With_Furniture.graph")$map
#names(gallery_w_furniture)
```
2) Process VGA Analysis : map without furniture
```{r}
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-local", radii = c("2,3,4,5,6,7"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("2"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("3"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("4"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("5"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("6"))
rdepthmap::VGA("Maps_Analysis/2_Without_Furniture.graph", vgaMode = "visibility-global", radii = c("7"))
gallery_wo_furniture = rdepthmap::getPointmapData("Maps_Analysis/2_Without_Furniture.graph")$map
#names(gallery_wo_furniture)
```
3) Process VGA Analysis : map for accessibility
```{r}
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-local", radii = c("2,3,4,5,6,7"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("2"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("3"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("4"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("5"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("6"))
rdepthmap::VGA("Maps_Analysis/3_Accessibility.graph", vgaMode = "visibility-global", radii = c("7"))
gallery_ac = rdepthmap::getPointmapData("Maps_Analysis/3_Accessibility.graph")$map
#names(gallery_wo_furniture)
```
List of metrics
```{r}
names(gallery_w_furniture)
```










# Load Danger Maps Data 

```{r message=FALSE, include=FALSE}
values = data.frame(gallery_w_furniture)
eth = readOGR("Maps_Ethnography/Ethnography.mif")
```


### Visualize Ethnography (Danger Map)
```{r}
eth.grid = as(eth,"SpatialPixelsDataFrame")
plot(eth.grid[, "Verified8k"])
plot(eth.grid[, "Verified9k"])
```


### Process Reddit Post with 8k Likes
```{r}
v_8k_df = data.frame(values)
v_8k_df$risk = eth$Verified8k
#write_csv(v_8k_df,"ouputdata/Verification8k.csv")

rsq_8k_df = purrr::map(v_8k_df[,-1],  ~lm(risk ~ .x, data = v_8k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_8k_df

f_rsq_8k_df  = rsq_8k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))
```
### Process Reddit Post with 9k Likes
```{r}
v_9k_df = data.frame(values)
v_9k_df$risk = eth$Verified9k
#write_csv(v_9k_df,"ouputdata/Verification9k.csv")

rsq_9k_df = purrr::map(v_9k_df[,-1],  ~lm(risk ~ .x, data = v_9k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_9k_df

f_rsq_9k_df  = rsq_9k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))
```




## Compare R.Squared Values
```{r}
p_8k <- ggplot(data=f_rsq_8k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_8k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

p_9k <- ggplot(data=f_rsq_9k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_9k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```
## Scatter different Visual Metrics
```{r warning=FALSE}
#ggplot(v_8k_df, aes(x=v_8k_df$risk, y=v_8k_df$Visual.Node.Count.R3)) + geom_point()
ggplot(v_8k_df, aes(x=v_8k_df$risk, y=v_8k_df$Visual.Relativised.Entropy.R6)) + geom_point(shape=1,size=5) 
```
## Visualize Actual on Map
```{r} 
#plot(gallery_w_furniture[, "Visual.Node.Count.R3"])
plot(gallery_w_furniture[, "Visual.Relativised.Entropy.R6"])
plot(gallery_w_furniture[, "Visual.Integration..P.value..R7"])
eth.grid = as(eth,"SpatialPixelsDataFrame")
plot(eth.grid[, "Verified8k"])
plot(eth.grid[, "Verified9k"])
```

# Load Kill Map Data

## Find Point Deaths
```{r warning=FALSE}
#Visualize Points
pd = readOGR("Maps_Ethnography/Deaths.mif")
plot(gallery_w_furniture[, "Visual.Integration..Tekl..R7"])
plot(pd, add = TRUE, col = "red")
```
Kernel Density Graph at 2500 RADIUS
```{r echo=TRUE}
ma = as(eth.grid[, "Verified8k"],"owin")
pp = as(pd,"ppp")
marks(pp) = NULL
Window(pp) = ma
K <- density(pp,sigma = 3000)  #more is more spread
plot(K, main=NULL, las=1)
contour(K, add=TRUE)
```
Convert Kernel Density Graph into SPDF to Visualize
```{r}
d = eth
d1 = cbind(d,K)
d1$x <- NULL
d1$y <- NULL
d1$optional <- NULL
d.grid = as(d1,"SpatialPixelsDataFrame")
plot(d.grid[, "value"])

pd = readOGR("Maps_Ethnography/Deaths.mif")
plot(pd, add = TRUE, col = "red")
```
Testing R Correlation
```{r}
v_0k_df = data.frame(values)
v_0k_df$risk = d1$value

rsq_0k_df = purrr::map(v_0k_df[,-1],  ~lm(risk ~ .x, data = v_0k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_0k_df

f_rsq_0k_df  = rsq_0k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))


p_0k <- ggplot(data=f_rsq_0k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_0k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```

## Using QGIS manipulated map
```{r}
#Visualize Heatmap Diagram
death = readOGR("Maps_Ethnography/Ethnography_2.mif")
death.grid = as(death,"SpatialPixelsDataFrame")
plot(death.grid[, "one"])
plot(death.grid[, "onehalf"])
plot(death.grid[, "two"])
plot(death.grid[, "twohalf"])
```
Check Distribution of Values (TOTAL 3000 points)
```{r include=FALSE}
library(purrr)
one = data.frame(death$one) %>% arrange(death.one) %>% filter(death.one>0) %>% mutate(col = row_number()) %>% .[c(2,1)]
onehalf = data.frame(death$onehalf) %>% arrange(death.onehalf) %>% filter(death.onehalf>0) %>% mutate(col = row_number()) %>% .[c(2,1)]
two = data.frame(death$two) %>% arrange(death.two) %>% filter(death.two>0) %>% mutate(col = row_number()) %>% .[c(2,1)]
twohalf = data.frame(death$twohalf) %>% arrange(death.twohalf) %>% filter(death.twohalf>0) %>% mutate(col = row_number()) %>% .[c(2,1)]
plot(one)
plot(onehalf)
plot(two)
plot(twohalf)
```
Correlation Test
```{r warning=FALSE, include=FALSE}
#one----------------------------------------------
v_1k_df = data.frame(values)
v_1k_df$risk = death$one
#write_csv(v_1k_df,"ouputdata/Verification1k.csv")

rsq_1k_df = purrr::map(v_1k_df[,-1],  ~lm(risk ~ .x, data = v_1k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_1k_df

f_rsq_1k_df  = rsq_1k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))

#onehalf----------------------------------------------

v_1_5k_df = data.frame(values)
v_1_5k_df$risk = death$onehalf
#write_csv(v_1_5k_df,"ouputdata/Verification1_5k.csv")

rsq_1_5k_df = purrr::map(v_1_5k_df[,-1],  ~lm(risk ~ .x, data = v_1_5k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_1_5k_df

f_rsq_1_5k_df  = rsq_1_5k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))

#two----------------------------------------------

v_2k_df = data.frame(values)
v_2k_df$risk = death$two
#write_csv(v_2k_df,"ouputdata/Verification2k.csv")

rsq_2k_df = purrr::map(v_2k_df[,-1],  ~lm(risk ~ .x, data = v_2k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_2k_df

f_rsq_2k_df  = rsq_2k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))

#twohalf----------------------------------------------

v_2_5k_df = data.frame(values)
v_2_5k_df$risk = death$twohalf
#write_csv(v_2_5k_df,"ouputdata/Verification2_5k.csv")

rsq_2_5k_df = purrr::map(v_2_5k_df[,-1],  ~lm(risk ~ .x, data = v_2_5k_df)) %>%
  purrr::map(broom::glance) %>%
  dplyr::bind_rows(.id = "variable") %>%
  dplyr::select(variable, r.squared)
#rsq_2_5k_df

f_rsq_2_5k_df  = rsq_2_5k_df %>% 
  .[-c(46, 47, 48, 49,50), ] %>% 
  arrange(desc(r.squared))

```

## Correlation Based on Heatmap
```{r}
p_1k <- ggplot(data=f_rsq_1k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_1k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

p_1_5k <- ggplot(data=f_rsq_1_5k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_1_5k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

p_2k <- ggplot(data=f_rsq_2k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_2k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

p_2_5k <- ggplot(data=f_rsq_2_5k_df, aes(x=variable, y=r.squared)) +
  geom_bar(stat="identity")
p_2_5k + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```


## Distribution of Data Points based on Visual Metrics
```{r, fig.width = 15, fig.asp = .62}
overlapsPD = over(pd,gallery_w_furniture)
overlapsPD.a <- scale(overlapsPD[,2:46],center=T,scale=T)
overlapsPD.b = as.data.frame(overlapsPD.a)
overlapsPD.b$Ref =overlapsPD$Ref
DAT.melt = melt(overlapsPD.b,id=c("Ref"))
ggplot(DAT.melt, aes(x=variable,y=value) ) + geom_point(shape=1,size=4)  + theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))
```


# General Map Visibility

## Compare visibility radius 2-7

### 1. map with furniture
```{r}
plot(gallery_w_furniture[, "Visual.Integration..HH..R2"])
plot(gallery_w_furniture[, "Visual.Integration..HH..R3"])
plot(gallery_w_furniture[, "Visual.Integration..HH..R7"])
```
### 2. map without furniture
```{r}
plot(gallery_wo_furniture[, "Visual.Integration..HH..R2"])
plot(gallery_wo_furniture[, "Visual.Integration..HH..R3"])
plot(gallery_wo_furniture[, "Visual.Integration..HH..R7"])
```
### 3. map accessibility
```{r}
plot(gallery_ac[, "Visual.Integration..HH..R2"])
plot(gallery_ac[, "Visual.Integration..HH..R3"])
plot(gallery_ac[, "Visual.Integration..HH..R7"])
```


### Compare with and without furniture

#### Visual Integration R7
```{r}
plot(gallery_w_furniture[, "Visual.Integration..HH..R7"])
plot(gallery_wo_furniture[, "Visual.Integration..HH..R7"])
```
#### Visual Entropy, R6
```{r}
plot(gallery_w_furniture[, "Visual.Relativised.Entropy.R6"])
plot(gallery_wo_furniture[, "Visual.Relativised.Entropy.R6"])
```







# Feature Visibility

## Visualizing Feature Spaces
```{r, fig.width=8}
#names(gallery_w_furniture)

feature_layer = readOGR("Maps_Analysis/Feature_Polygon2.mif", verbose = FALSE)
plot(gallery_w_furniture[, "Visual.Integration..HH..R7"])
plot(feature_layer, add = TRUE, col = "red")
```
## Complete Feature List
```{r}
overlapsDF = over(gallery_w_furniture[, "Visual.Integration..HH..R7"], feature_layer)
visi_df = aggregate(gallery_w_furniture$Visual.Integration..HH..R7, by = list(overlapsDF$Layer,overlapsDF$Text,overlapsDF$LengthTask,overlapsDF$VisualTask), mean, areaWeighted = TRUE)
vif = visi_df %>% arrange(Group.4)
write.csv(vif,"ouputdata/visualintegration_features.csv")
```




## Comparisons by Averaging

@Types_of_Task
```{r message=FALSE}
visi_df %>%
        	group_by(Group.1) %>%
        	summarise(V_r7 = mean(x))
```
@Length_of_Task
```{r message=FALSE}
visi_df %>%
          filter(Group.1 == "tasks" | Group.1 == "hotpink" | Group.1 == "lightpink") %>%
        	group_by(Group.3) %>%
        	summarise(V_r7 = mean(x))
```
@Visual_Type_of_Task
```{r message=FALSE}
visi_df %>%
          filter(Group.1 == "tasks" | Group.1 == "hotpink" | Group.1 == "lightpink") %>% 
        	group_by(Group.4) %>%
        	summarise(V_r7 = mean(x))
```

Plot of Visibility for Each Tasks
```{r}
task_all_df = visi_df %>% filter(Group.1 == "tasks") %>% .[c(3,5)] 
task_time_df = visi_df %>% filter(Group.1 == "tasks") %>% .[c(3,5)]
task_visual_df = visi_df %>% filter(Group.1 == "tasks") %>% .[c(4,5)]

task_all_df$Group.3 = "All"
taskall = melt(task_all_df,id=c("Group.3"))
tasklength = melt(task_time_df,id=c("Group.3"))
taskvisual = melt(task_visual_df,id=c("Group.4")) %>% filter(Group.4 == 'Visual')
names(taskvisual)[1] <- "Group.3"

task_df = rbind(taskall,tasklength,taskvisual)
ggplot(task_df, aes(x=Group.3,y=value, color = Group.3) ) + geom_boxplot(shape=2,size=1)  + theme(axis.text.x = element_text(angle = 0, vjust = 0.5, hjust=0.5)) + labs(x="")
```







## Feature Tables
@Types_of_Task
```{r}
visi_df %>% filter(Group.1 == "tasks")
visi_df %>% filter(Group.1 == "sabotage")
visi_df %>% filter(Group.1 == "surveillance")
visi_df %>% filter(Group.1 == "vents")
```
@Length_of_Task
```{r}
visi_df %>% filter(Group.3 == "Long") %>% select(-one_of(c("Group.1","Group.2")))
visi_df %>% filter(Group.3 == "Common") %>% select(-one_of(c("Group.1","Group.2")))
visi_df %>% filter(Group.3 == "Short") %>% select(-one_of(c("Group.1","Group.2")))
```
@Visual_Type_of_Task
```{r}
visi_df %>% filter(Group.4 == "Visual") %>% select(-one_of(c("Group.1")))
visi_df %>% filter(Group.1 == "tasks" | Group.1 == "hotpink" | Group.1 == "lightpink") %>% filter(Group.4 != "Visual") %>% select(-one_of(c("Group.1")))
```