library(ggplot2)

# Data Source: https://www.kaggle.com/datasets/samuelotiattakorah/agriculture-crop-yield?resource=download

df <- read.csv("crop_yield.csv", header = TRUE, sep = ",")
df <- df[df$Crop == "Soybean", ]

# Single Boxplot 

ggplot(df, aes(x = "", y = Yield_tons_per_hectare)) +
  geom_boxplot(fill = "#D44803", width = 0.3) +  # make box narrower
  labs(y = "Yield", x = NULL, title = "Boxplots with PikBioStat: Soyabean Yield") +
  theme_minimal(base_size = 14) +
  theme(
    panel.grid = element_blank(),   # remove gridlines
    axis.text.x = element_blank(),  # hide x-axis labels
    axis.ticks.x = element_blank(), # hide x-axis ticks
    axis.text.y = element_text(size = 14), # bigger y-axis labels
    plot.title = element_text(hjust = 0.5) # center title
  )


# Group Comparisons 

ggplot(df, aes(x = Fertilizer_Used, y = Yield_tons_per_hectare)) +
  geom_boxplot(fill = "#D44803", width = 0.5) +  # multiple boxes by Soil_Type
  labs(y = "Yield (tons/ha)", x = "Fertilizer Used", 
       title = "Boxplots with PikBioStat: Soybean Yield by Fertilizer Used") +
  theme_minimal(base_size = 14) +
  theme(
    panel.grid = element_blank(),   
    axis.text.x = element_text(size = 12, angle = 45, hjust = 1), # readable labels
    axis.text.y = element_text(size = 14), 
    axis.title.x = element_text(size = 14, face = "bold"),
    axis.title.y = element_text(size = 14, face = "bold"),
    plot.title = element_text(hjust = 0.5, size = 16, face = "bold")
  )