# Titanic: skrypt startowy do zajec z ggplot2
# Uruchom ten plik przed pisaniem wykresow na zywo.

library(tidyverse)
library(scales)

theme_set(theme_minimal(base_size = 14))

if (requireNamespace("rstudioapi", quietly = TRUE) && rstudioapi::isAvailable()) {
  script_path <- rstudioapi::getSourceEditorContext()$path
  if (nzchar(script_path)) {
    setwd(dirname(script_path))
  }
}

dir.create("data", showWarnings = FALSE)
data_path <- "data/train.csv"

if (!file.exists(data_path)) {
  url <- "https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv"
  download.file(url, data_path, mode = "wb", quiet = TRUE)
}

train <- read_csv(data_path, show_col_types = FALSE)

message("Katalog pracy: ", getwd())
message("Plik danych: ", normalizePath(data_path))

titanic <- train %>%
  mutate(
    Survived = factor(
      Survived,
      levels = c(0, 1),
      labels = c("Nie przeżył", "Przeżył")
    ),
    Pclass = factor(
      Pclass,
      levels = c(1, 2, 3),
      labels = c("1 klasa", "2 klasa", "3 klasa")
    ),
    Sex = factor(
      Sex,
      levels = c("female", "male"),
      labels = c("kobieta", "mężczyzna")
    ),
    Embarked = factor(
      Embarked,
      levels = c("C", "Q", "S"),
      labels = c("Cherbourg", "Queenstown", "Southampton")
    ),
    Title = gsub("(.*, )|(\\..*)", "", Name),
    Title = case_when(
      Title %in% c("Mlle", "Ms") ~ "Miss",
      Title == "Mme" ~ "Mrs",
      Title %in% c("Mr", "Miss", "Mrs", "Master") ~ Title,
      TRUE ~ "Inny"
    ),
    Title = factor(Title, levels = c("Mr", "Miss", "Mrs", "Master", "Inny")),
    FamilySize = SibSp + Parch + 1,
    FamilyGroup = case_when(
      FamilySize == 1 ~ "samodzielnie",
      FamilySize <= 4 ~ "mała rodzina",
      TRUE ~ "duża rodzina"
    ),
    FamilyGroup = factor(
      FamilyGroup,
      levels = c("samodzielnie", "mała rodzina", "duża rodzina")
    ),
    AgeGroup = case_when(
      is.na(Age) ~ "brak wieku",
      Age < 12 ~ "dziecko",
      Age < 18 ~ "nastolatek",
      Age < 40 ~ "młody dorosły",
      Age < 60 ~ "dorosły",
      TRUE ~ "senior"
    ),
    AgeGroup = factor(
      AgeGroup,
      levels = c(
        "dziecko",
        "nastolatek",
        "młody dorosły",
        "dorosły",
        "senior",
        "brak wieku"
      )
    )
  )

survival_palette <- c("Nie przeżył" = "#d94f45", "Przeżył" = "#2f6f8f")
class_palette <- c("1 klasa" = "#2f6f8f", "2 klasa" = "#f2a541", "3 klasa" = "#7a869a")

missing_data <- tibble(
  Variable = names(train),
  Missing = colSums(is.na(train))
) %>%
  filter(Missing > 0) %>%
  arrange(desc(Missing))

glimpse(titanic)
missing_data

# Od tego miejsca piszemy wykresy: