Histogram for each season for indicated variable

UnivariateAnalysis.R

@@ -3,8 +3,27 @@ setwd("/home/mariogt/TFGs/Estadistica/")
 #####################################
 #             FUNCIONES             #
 #####################################
-# Returns the distribution of a variable from 3 seasons with a bar plot.
-plotFor3Seasons <- function(data, variable) {
+# Histograms for all seasons for the variable entered by parameter.
+barpFor3Season <- function(data, variable) {
+  par(mfrow = c(1, 3))
+  dfs <- list()
+  maxPInPos <- 0
+  for (i in 1:3) {
+    season <- switch(i, ".1718", ".1819", ".1920")
+    var <- select(data, contains(paste(variable, season, sep = "")))
+    cat(names(var))
+    x <- as.numeric(var[,1])
+    h <- hist(x, 
+                      main = paste("Histograma para", titleForYAxis(variable), "(", season, ")", sep = " "), 
+                      col = "darkolivegreen3", 
+                      border = "mediumorchid4",
+                      breaks = seq(from = 0, to = max(x) + 5, by = 5))
+    text(seq(2.5, max(x) + 2.5, 5), h$counts, labels = h$counts, adj = c(0.5, -0.5))
+  }
+}
+
+# Returns the count of a variable from 3 seasons for each position with a bar plot.
+plotFor3SeasonsForPos <- function(data, variable) {
   par(mfrow = c(1, 3))
   dfs <- list()
   maxPInPos <- 0
@@ -25,11 +44,13 @@ plotFor3Seasons <- function(data, variable) {
       position = c("Portero", "Defensa", "Mediocentro", "Delantero"),
       numberOf = c(numGK, numDF, numMF, numFW)
     )
+    cat(df$numberOf)
     dfs[[i]] <- df
     if(maxPInPos < max(df$numberOf)) {
       maxPInPos <- max(df$numberOf)
     }
   }
+  
   for (i in 1:3) {
     season <- switch(i, ".1718", ".1819", ".1920")
     bp <- barplot(height = dfs[[i]]$numberOf, names = dfs[[i]]$position, 
@@ -53,14 +74,14 @@ univariantAnalysis <- function(data, variable, ids) {
   
   bp <- ggplot(dl, aes(x = factor(season), y = variable)) +
     geom_boxplot(alpha = 0.5, fill="darkblue", color="black", outlier.color="red")+
-    labs(title="Boxplot GCA por temporada", x = "Temporada", y = "Acciones de creación de gol")
+    labs(title="Boxplot por temporada", x = "Temporada", y = titleForYAxis(variable))
   
   g <- group_by(dl, season)
   g2 <- summarise(g, media = round(mean(variable), 1))
   gp <- ggplot(g2, aes(x = season, media, group = 1))+
     geom_point(alpha=0.5, color = "blue", size=3) +
     geom_line(color = "red", size = 1) +
-    labs(title = "Promedio de GCA por temporada", x = "Temporada", y = "Acciones de creación de gol")
+    labs(title = "Promedio por temporada", x = "Temporada", y = titleForYAxis(variable))
   
   figure <- ggarrange(bp, gp,
                       labels = c("A", "B"),
@@ -126,85 +147,52 @@ ids <- t(ids)
       ##################################
       # Goles y asistencias por 90 min #
       ##################################
+      univariantAnalysis(train, "goals_assists_per90", ids)
       ######
       # xG #
       ######
+      univariantAnalysis(train, "xg", ids)
       ########
       # npxG #
       ########
+      univariantAnalysis(train, "npxg", ids)
       ######
       # xA #
       ######
+      univariantAnalysis(train, "xa", ids)
       ##################################
       # Porcentaje de tiros a porteria #
       ##################################
+      
       #######################
       # Porcentaje de pases #
       #######################
-
-passes_pct <- train %>% select(contains("passes_pct."))
-passes_pct <- cbind(ids, passes_pct)
-
-
-data_long = gather(passes_pct, season, passes, passes_pct.1718:passes_pct.1920, convert = TRUE, factor_key=TRUE)
-data_long
-data_long[, 3] <- as.numeric(data_long[, 3])
-
-
-ggplot(data_long, aes(x = factor(season), y = passes)) +
-  geom_boxplot(alpha = 0.5, fill="darkblue", color="black", outlier.color="red")+
-  labs(title="Boxplot susceptibilidad por tiempo y entrenamiento", x="Temporada",
-       y = "Porcentaje de pases exitosos")
+      univariantAnalysis(train, "passes_pct", ids)
       ##################################
       # Distancia total mediante pases #
       ##################################
+      univariantAnalysis(train, "passes_total_distance", ids)
       #############################
       # Pases que derivan en tiro #
       #############################
+      plotFor3Seasons(train, "assisted_shots")
+      univariantAnalysis(train, "assisted_shots", ids)
       ###################################
       # Cambios de orientación de juego #
       ###################################
+      univariantAnalysis(train, "passes_switches", ids)
       ###################
       # SCA passes dead #
       ###################
-sca_passes_dead <- train %>% select(contains("sca_passes_dead."))
-sca_passes_dead <- cbind(ids, sca_passes_dead)
-
-data_long = gather(sca_passes_dead, season, sca_passes_dead, sca_passes_dead.1718:sca_passes_dead.1920, convert = TRUE, factor_key=TRUE)
-data_long
-data_long[, 3] <- as.numeric(data_long[, 3])
-
-ggplot(data_long, aes(x = factor(season), y = sca_passes_dead)) +
-  geom_boxplot(alpha = 0.5, fill="darkblue", color="black", outlier.color="red")+
-  labs(title="Boxplot SCA por temporada", x="Temporada",
-       y = "Acciones de creación de tiro")
-
+      univariantAnalysis(train, "sca_passes_dead", ids)
       ###################
       # GCA passes dead #
       ###################
-
-
-univariantAnalysis(train, "passes_intercepted", ids)
-gca_passes_dead <- train %>% select(contains("gca_passes_dead."))
-gca_passes_dead <- cbind(ids, gca_passes_dead)
-par(mfrow = c(2, 1))
-data_long = gather(gca_passes_dead, season, gca_passes_dead, gca_passes_dead.1718:gca_passes_dead.1920, factor_key=TRUE)
-data_long
-data_long[, 3] <- as.numeric(data_long[, 3])
-
-ggplot(data_long, aes(x = factor(season), y = gca_passes_dead)) +
-  geom_boxplot(alpha = 0.5, fill="darkblue", color="black", outlier.color="red")+
-  labs(title="Boxplot GCA por temporada", x = "Temporada", y = "Acciones de creación de gol")
-
-g <- group_by(data_long, season)
-g2 <- summarise(g, media = round(mean(gca_passes_dead), 1))
-ggplot(g2, aes(x = season, media, group = 1))+
-  geom_point(alpha=0.5, color = "blue", size=3) +
-  geom_line(color = "red", size = 1) +
-  labs(title = "Promedio de GCA por temporada", x = "Temporada", y = "Acciones de creación de gol")
+      univariantAnalysis(train, "gca_passes_dead", ids)
       #######################
       # Pases interceptados #
       #######################
+      univariantAnalysis(train, "passes_intercepted", ids)
       ####################################
       # Porcentaje de presiones exitosas #
       ####################################
@@ -214,16 +202,21 @@ ggplot(g2, aes(x = season, media, group = 1))+
       ####################
       # Faltas cometidas #
       ####################
+      univariantAnalysis(train, "fouls", ids)
       #######################################
       # Porcentaje de juegos aereos ganados #
       #######################################
       ######################
       # Penalties atajados #
       ######################
+      univariantAnalysis(train, "pens_att", ids)
       ##############################
       # Goles en contra por 90 min #
       ##############################
+      univariantAnalysis(train, "goals_against_per90_gk", ids)
       ###################################
       # Porcentaje de ocasiones paradas #
       ###################################
+      univariantAnalysis(train, "")
+
 laLigaPlayers$foot <- as.factor(laLigaPlayers$foot)