library(googleVis)
ls("package:googleVis")

# Diagrama de dispersão.
p <- gvisScatterChart(
    data = cars[, c("speed", "dist")],
    options = list(legend = "none",
                   lineWidth = 0,
                   pointSize = 1,
                   title = "Distância x Velocidade",
                   vAxis = "{title: 'Distância'}",
                   hAxis = "{title: 'Velocidade'}",
                   width = 700,
                   height = 500))

# Exibe o código fonte completo.
p

# getOption("gvis.tags")
# Configure o options() e use `results = "asis"` no chunk.
options(gvis.plot.tag = "chart")

# Para não exibir metadados.
# p$html$footer <- NULL
# p$html$caption <- NULL

# Exibe um fragmento HTML.
plot(p)

br <- data.frame(
    Estado = c("São Paulo", "Minas Gerais", "Rio de Janeiro", "Bahia",
               "Rio Grande do Sul", "Paraná", "Pernambuco", "Ceará",
               "Pará", "Maranhão", "Santa Catarina", "Goiás", "Paraíba",
               "Amazonas", "Espírito Santo", "Rio Grande do Norte",
               "Alagoas", "Mato Grosso", "Piauí", "Distrito Federal",
               "Mato Grosso do Sul", "Sergipe", "Rondônia", "Tocantins",
               "Acre", "Amapá", "Roraima"),
    População = c(44396484L, 20869101L, 16550024L, 15203934L, 11247972L,
                  11163018L, 9345173L, 8904459L, 8175113L, 6904241L,
                  6819190L, 6610681L, 3972202L, 3938336L, 3929911L,
                  3442175L, 3340932L, 3270973L, 3204028L, 2914830L,
                  2651235L, 2242937L, 1768204L, 1515126L, 803513L,
                  766679L, 505665L),
    stringsAsFactors = FALSE)

br$População <- log10(br$População)
str(br)

breaks <- seq(floor(min(br$População)),
              ceiling(max(br$População)),
              by = 0.5)

library(RColorBrewer)
pal <- brewer.pal(n = length(breaks), name = "Blues")

# Dicionário para os intervalos de classe e cores associadas.
cl <- sprintf("{values:[%s],\n colors:[%s]}",
              paste0(sprintf("'%0.1f'", breaks), collapse = ", "),
              paste0(sprintf("'%s'", pal), collapse = ", "))
cat(cl)

map <- gvisGeoChart(
    data = br,
    locationvar = "Estado",
    colorvar = "População",
    options = list(title = "lala",
                   region = "BR",
                   displayMode = "regions",
                   resolution = "provinces",
                   colorAxis = cl,
                   width = 700,
                   height = 500))
plot(map)

br <- data.frame(
    capital = c("Aracaju - SE", "Belém - PA", "Belo Horizonte - MG",
                "Boa Vista - RR", "Brasília - DF", "Campo Grande - MS",
                "Cuiabá - MT", "Curitiba - PR", "Florianópolis - SC",
                "Fortaleza - CE", "Goiânia - GO", "João Pessoa - PB",
                "Macapá - AP", "Maceió - AL", "Manaus - AM",
                "Natal - RN", "Palmas - TO", "Porto Alegre - RS",
                "Porto Velho - RO", "Recife - PE", "Rio Branco - AC",
                "Rio de Janeiro - RJ", "Salvador - BA", "São Luís - MA",
                "São Paulo - SP", "Teresina - PI", "Vitória - ES"),
    lat = c(-10.911, -1.456, -19.921, 2.82, -15.78, -20.443, -15.596,
            -25.428, -27.597, -3.717, -16.679, -7.115, 0.039, -9.666,
            -3.102, -5.795, -10.213, -30.033, -8.762, -8.054, -9.975,
            -22.903, -12.971, -2.53, -23.548, -5.089, -20.319),
    lon = -c(37.072, 48.504, 43.938, 60.673, 47.93, 54.646, 56.097,
             49.273, 48.549, 38.543, 49.254, 34.863, 51.066, 35.735,
             60.025, 35.209, 48.36, 51.23, 63.904, 34.881, 67.81,
             43.208, 38.511, 44.303, 46.636, 42.802, 40.338),
    stringsAsFactors = FALSE)
br$latlon <- with(br, paste(lat, lon, sep = ":"))
str(br)

map <- gvisMap(data = br,
               locationvar = "latlon",
               tipvar = "capital",
               options = list(showTip = TRUE,
                              showLine = TRUE,
                              enableScrollWheel = TRUE,
                              mapType = "terrain",
                              useMapTypeControl = TRUE,
                              width = "700px",
                              height = "500px"))
plot(map)

library("highcharter")

data(mpg, package = "ggplot2")

hc <- hchart(mpg,
             type = "scatter",
             mapping = hcaes(x = "displ", y = "hwy", group = "class"))
hc

data(diamonds, package = "ggplot2")

hc <- hchart(diamonds$price,
             color = "#B71C1C",
             name = "Price") %>%
    hc_title(text = "You can zoom me")
hc

library(plotly)
head(ls("package:plotly"), n = 20)

gg <-
    ggplot(data = cars,
           mapping = aes(x = speed, y = dist)) +
    geom_point() +
    geom_smooth() +
    xlab("Velocidade") +
    ylab("Distância")

ggplotly(gg, width = 600, height = 400)

data(iris)

irisw <- reshape2::melt(iris, id.vars = "Species")

gg <-
    ggplot(data = irisw,
           mapping = aes(fill = Species, x = value)) +
    geom_density(alpha = 0.3) +
    facet_wrap(facets = ~variable, scales = "free")

ggplotly(gg, width = 800, height = 600)

gg <-
    ggplot(irisw, aes(x = Species, y = value)) +
    geom_boxplot() +
    facet_wrap(facets = ~variable, scales = "free")

ggplotly(gg, width = 800, height = 600)

library(leaflet)

# NOTE: As coordenadas do Google não batem exato com as coordenadas do
# leaflet.

coords <- data.frame(lat = c(-25.4503, -25.4508577),
                     lng = c(-49.2306, -49.2314582),
                     text = c("<strong>LEG</strong><br/>2º Andar, Sala 232",
                              "<strong>PET Estatística UFPR</strong><br/>PC-09"))

# names(providers)

leaflet() %>%
    setView(lat = mean(coords$lat),
            lng = mean(coords$lng),
            zoom = 18) %>%
    addProviderTiles("OpenStreetMap") %>%
    addPopups(lng = coords$lng,
              lat = coords$lat,
              coords$text,
              options = popupOptions(closeButton = FALSE))

#-----------------------------------------------------------------------
# Plotando no mapa de Curitiba.

url <- "http://leg.ufpr.br/~walmes/data/coberturas-venda-cwb-26Jan2018.txt"
tb <- read.table(url,
                 header = TRUE,
                 quote = "",
                 sep = "\t",
                 comment.char = "",
                 skip = 5,
                 stringsAsFactors = FALSE)
str(tb)

tb2 <- subset(tb, is.finite(lat), select = c("lat", "lon", "price"))
str(tb2)

leaflet(data = tb2) %>%
    addProviderTiles("OpenStreetMap") %>%
    addCircleMarkers(lng = ~lon,
                     lat = ~lat,
                     stroke = FALSE,
                     fillOpacity = 0.5,
                     radius = 5)

library(geojsonio)

# ATTENTION: não usar `what = "sp"` na chamada da `geojson_read()` para
# poder usar `addGeoJSON()`. Se o objeto for de classe baseada em `sp`
# tem-se que usar `addPolygons()`.

mp <- geojson_read("http://leg.ufpr.br/~walmes/data/bairros-cwb.kml",
                   method = "local")
class(mp)

leaflet() %>%
    setView(lat = -25.4842287,
            lng = -49.370973,
            zoom = 11) %>%
    # addTiles() %>%
    # addProviderTiles("OpenStreetMap") %>%
    addGeoJSON(geojson = mp,
               weight = 2,
               color = "green",
               fillColor = "yellow") %>%
    addCircleMarkers(data = tb2, lng = ~lon,
                     lat = ~lat,
                     stroke = FALSE,
                     fillOpacity = 0.5,
                     radius = 5)

library(animation)

da <- data.frame(x = seq(0, 2, by = 0.1))
da$y <- c(5.5, 7.2, 10.4, 7.4, 7.2, 9.1, 15.9, 12.7, 11.3, 14.4, 14.8,
          17.1, 31.4, 17.5, 27.6, 19.6, 27.1, 21, 33.8, 30.2, 45.1)

x_grid <- seq(0, 2, lenght.out = 1000)
b_grid <- seq(0, 2, by = 0.05)
b_grid <- b_grid[-c(1, length(grid))]
desvios <- rep(NA, length(b_grid))

# Demostração de uma regressão segmentada com diferentes pontos de
# quebra.
anim_fun <- function() {
    xlim <- extendrange(da$x, f = 0.025)
    for (i in 1:length(b_grid)) {
        #------------------------------------
        # Ajuste do modelo com valor do ponto de quebra fixo.
        b <- b_grid[i]
        m0 <- lm(y ~ x + I(pmax(x - b, 0)),
                 data = da)
        beta <- coefficients(m0)
        desvios[i] <- sum(residuals(m0)^2)
        cols <- c(rep(3, sum(da$x < b_grid[i])),
                  rep(4, sum(da$x > b_grid[i] | da$x == b_grid[i])))
        #------------------------------------
        # Gráfico.
        par(mfrow = c(1, 2), mar = c(3, 3, 1, 1))
        #------------------------------------
        # Gráfico 1.
        plot(y ~ x, xlim = xlim, data = da, ann = FALSE, pch = 19,
             col = cols)
        mtext(text = "x", side = 1, line = 2)
        mtext(text = "y", side = 2, line = 2)
        curve(beta[1] + beta[2] * x,
              from = 0, to = b, col = 3, add = TRUE)
        curve(beta[1] + beta[2] * b + (beta[3] + beta[2]) * (x - b),
              from = b, to = 2, col = 4, lwd = 2, add = TRUE)
        abline(v = b)
        #------------------------------------
        # Gráfico 2.
        plot(b_grid, xlim = xlim, ylim = c(300, 500), desvios,
             type = "o", pch = 1, xlab = "Pontos de Quebra")
        mtext(text = "x", side = 1, line = 2)
        mtext(text = "SQE", side = 2, line = 2)
        abline(v = b)
        ani.pause(interval = 0.1)
    }
}

anim_fun()

# Salva em HTML.
wname <- "piecewise.html"
saveHTML(anim_fun(),
         interval = 0.1,
         autobrowse = FALSE,
         htmlfile = wname,
         img.name = sub(x = wname, "\\.html$", ""),
         imgdir = sub(x = wname, "\\.html$", ""),
         verbose = FALSE,
         ani.height = 300,
         ani.width = 600)

library(knitr)
library(rgl)
knit_hooks$set(webgl = hook_webgl)

library(rgl)

str(volcano)

persp3d(volcano, col = "green")

da <- expand.grid(r = 1:2, C = -1:1, B = -1:1, A = -1:1,
                  KEEP.OUT.ATTRS = FALSE)
da$y <- c(-35, -25, 110, 75, 4, 5, -45, -60, -10, 30, -40, -30, -40, 15,
          80, 54, 31, 36, 17, 24, 55, 120, -23, -5, -65, -58, -55, -44,
          -64, -62, 20, 4, 110, 44, -20, -31, -39, -35, 90, 113, -30,
          -55, -55, -67, -28, -26, -61, -52, 15, -30, 110, 135, 54, 4)
str(da)

# Modelo reduzido.
m0 <- lm(y ~ (poly(A, 2) + poly(B, 2) + poly(C, 2))^2,
         data = da)

pred <- list(B = seq(-1.2, 1.2, by = 0.05),
             C = seq(-1.2, 1.2, by = 0.05))
pred$fxy <- outer(pred$B,
                  pred$C,
                  FUN = function(x, y) {
                      predict(m0,
                              newdata = data.frame(B = x,
                                                   C = y,
                                                   A = 0))
                  })

persp3d(x = pred$B,
        y = pred$C,
        z = pred$fxy,
        zlim = range(da$y),
        alpha = 0.7,
        color = "tomato")
with(subset(da, A == 0), {
    plot3d(x = B,
           y = C,
           z = y,
           type = "s",
           size = 1)
})