knitr::opts_chunk$set(
    echo = TRUE, 
    tidy.opts = list(width.cutoff = 65),
    tidy = TRUE)

set.seed(12314159)
library(loon)
data(minority)

imageDirectory <- "/Users/rwoldford/Documents/Research/talks/MontrealRtutorial2018/Slides/img"
dataDirectory <- "/Users/rwoldford/Documents/Research/talks/MontrealRtutorial2018/Slides/data"
path_concat <- function(path1, path2, sep="/") {paste(path1, path2, sep = sep)}
library(loon)
data(minority)
names(minority)[1:14]
subset <- t(minority[c("Victoria", "Ottawa", "St. John\'s"),][,1:(length(names(minority))-4)])
knitr::kable(subset)
minorityPercent <- data.frame(100*minority[, -c(11, 15:18)]
                              / minority[, "Total.population"])
#  Take the opportunity to shorten the Statistics Canada 
#  names of the minorities as well:
names(minorityPercent) <- c("% Arabic",  "% Black", "% Chinese",  
                            "% Filipino", "% Japanese",
                            "% Korean",  "% Latino", "% Multiple",
                            "% S. Asian",  "% SE Asian",  "% Other",
                            "% Minority", "% W Asian")
# The proportion of the city population that is a "visible minority"
l_hist(minorityPercent$"% Minority",
       title = "% Minority in 33 Canadian cities",
       xlabel = "percent of population",
       linkingGroup = "minority",
       yshows = "frequency",
       showBinHandle = FALSE,
       showGuides = TRUE,
       showScales = TRUE
)
# Order minorities by decreasing maximum percent
groups <- names(minorityPercent)[-c(12, 14, 15)]
bySize <- names(sort(apply(minorityPercent[,groups ], 2, max),
                       decreasing = TRUE) )
# And now look at the three largest
for (x in bySize[1:3]) {
  l_hist(minorityPercent[,x],
         title = x,
         xlabel = "percent of population",
         linkingGroup = "minority",
         showScales = TRUE,
         showGuides = TRUE,
         showStackedColors = TRUE,
         color = "grey"
         )
}
# get a map:
library(maps)
canada_map <- map("world", "Canada", plot=FALSE, fill=TRUE)
# plot the cities
cities_map <- with(minority,
                   l_plot(long, lat,
                          xlabel = "longitude", ylabel="latitude", showLabels = TRUE,
                          linkingGroup = "minority",
                          itemLabel = rownames(minority), showItemLabels = TRUE)
)

# Layer in the map of Canada
landcol <- "cornsilk"
canada_layer <- l_layer(widget = cities_map, 
                        x = canada_map, 
                        label = "Canada",
                        color = landcol,
                        index = "end")
# Rescale the plot to the size of the map
l_scaleto_layer(cities_map, canada_layer)

# Could also add the text of the city names to the map as glyphs
textGlyphs <- l_glyph_add_text(cities_map, text = row.names(minority),
                               label = "city names")
minPercentPairs <- l_pairs(data = minorityPercent,
                           size = 2, glyph = "ocircle",
                           linkingGroup = "minority",
                           itemLabel = rownames(minority),
                           showItemLabels = TRUE)
# a compound loon object consisting of many plots
names(minPercentPairs) [1:5]
minPercentPairs$x2y1
minPercentPairs$x2y1["glyph"] <- "otriangle"
# Back to the raw counts 
minority_only <- minority[, c(
    "Southeast.Asian", "Chinese", "Japanese", "South.Asian", 
    "Visible.minority.not.included.elsewhere",
    "Black", "Multiple.visible.minority", "Filipino",
    "Arab", "Korean", "Latin.American")]
# Again, shorten the variable names
names(minority_only) <- c(
    "SE.Asian", "Chinese", "Japanese", "S.Asian", "Other",
    "Black", "Multiple", "Filipino", "Arab", "Korean", "Latino"
)
s <- l_serialaxes(data = minority_only, 
                  axesLayout = "parallel",   # Note choice of parallel
                  linkingGroup = "minority",
                  sequence = names(minority_only),
                  showGuides = FALSE,
                  linewidth = 2,
                  scaling = "data",
                  showArea = FALSE,
                  itemLabel = rownames(minority),  
                  showItemLabels = TRUE)
s_glyphs <- l_glyph_add_serialaxes(cities_map, 
                                   data = minority_only,
                                   sequence = names(minority_only),
                                   scaling = "variable",
                                   showArea = TRUE,
                                   label = "serial_axes")
cities_map['glyph'] <- s_glyphs
barplot_imgs <- l_make_glyphs(data = lapply(1:nrow(minority_only), 
                                            FUN = function(i) minority_only[i,]), 
                              # draw a coloured bar plot
                              draw_fun = function(m) {
                                  par(mar=c(1,1,1,1)*.5)
                                  mat <- as.numeric(m/max(m))
                                  barplot(height = mat,
                                          beside = FALSE,
                                          ylim = c(0,1),
                                          col = rainbow(length(mat)),
                                          axes= FALSE,
                                          axisnames=FALSE)}, 
                              width=120, 
                              height=120)
# Can view the images using loon's viewer
img_vwr <- l_imageviewer(barplot_imgs)
# add the glyphs  
barplot_glyphs <- l_glyph_add_image(cities_map, barplot_imgs, "barplot")
cities_map['glyph'] <- barplot_glyphs
library(RnavGraphImageData) 
data(frey)
# Here are the images
frey.imgs <- l_image_import_array(frey, 28,20,
                                  img_in_row = FALSE, rotate = 90)
frey_imgs_vwr <- l_imageviewer(frey.imgs)
data <- t(frey)  # images were in columns
# get the principal components
frey_pc <- prcomp(data)
plot(frey_pc$sdev, type = "b", pch = 19, col = "blue", main = "Principal component scree", ylab = "singular values")
abline(v=6, col = "firebrick", lty = 2)
frey_pc_coords <- frey_pc$x[, 1:6]
# create a navigation graph
frey_pc_nav  <- l_navgraph(frey_pc_coords, linkingGroup="frey")
frey_glyphs  <- l_glyph_add_image(frey_pc_nav$plot, 
                                  images=frey.imgs, 
                                  label="frey faces")
# l_navigator_walk_path(frey_pc_nav$navigator, path = c("PC1:PC2", "PC2:PC3"))
library(grid)
grid.newpage()
grid.loon(frey_pc_nav)
## library(dimRed) # dimension reduction package
## # Number of neighbours
## k_nbhrs <- 12
## # Target dimension
## n_dims <- 20
## # Dimension reduction via LLE ... this may take a while
## lle20 <-  embed(data, "LLE", knn = k_nbhrs, ndim = n_dims)
## lle20data <- lle20@data@data
## l_serialaxes(lle20data, linkingGroup = "frey")
## # Get a set of "closures" which will calculate scagnostics.
## lle_scags2d <- scagnostics2d(lle20data)
## # Now a scatterplot matrix of THE PLOTS
## lle_nav <- l_ng_plots(measures=lle_scags2d, linkingGroup="frey", size = 1)
## 
## lle_gl <- l_glyph_add_image(lle_nav$plot,
##                             images=frey.imgs,
##                             label="frey faces")
## 
## lle_nav$plot["linkingGroup"] <- "frey"
## lle_nav$plots$x2y1["size"] <- 1