knitr::opts_chunk$set(echo = TRUE)

# Import the PAC learning algorithm package

library(PAC)

Vector containing a set of apples

# Defines a vector named 'fruits' for simplification all fruits are apples.
fruits <- rep("Apple", 20)

Randomly assign charisteristic to the apple color, size and sweet not sweet.

#Create a vector names 'colors' by repeating a sequence of colors for each apple in the vector 'fruits'

colors <- sample(c("Red", "Yellow", "Green", "Purple"), length(fruits), replace = TRUE)
sizes <- sample(c("Small", "Medium", "Large"), length(fruits), replace = TRUE)
sweetness <- sample(c("Sweet", "Not sweet"), length(fruits), replace = TRUE)

Create the dataframe named ‘fruits_df by combining the ’fruits’, ‘colors’, sizes’, and ’sweetness into a vector column

fruits_df <- data.frame(Fruit = fruits, Color = colors, Size = sizes, Sweetness = sweetness)
fruits_df

##    Fruit  Color   Size Sweetness
## 1  Apple Yellow  Large     Sweet
## 2  Apple Purple  Small Not sweet
## 3  Apple  Green Medium Not sweet
## 4  Apple    Red  Small     Sweet
## 5  Apple Yellow  Small     Sweet
## 6  Apple  Green  Small Not sweet
## 7  Apple    Red  Large     Sweet
## 8  Apple Yellow  Large Not sweet
## 9  Apple  Green  Large Not sweet
## 10 Apple  Green  Large     Sweet
## 11 Apple Yellow  Small Not sweet
## 12 Apple  Green  Large     Sweet
## 13 Apple  Green  Small     Sweet
## 14 Apple  Green  Large Not sweet
## 15 Apple Purple Medium     Sweet
## 16 Apple    Red Medium Not sweet
## 17 Apple    Red Medium Not sweet
## 18 Apple    Red Medium     Sweet
## 19 Apple  Green  Large     Sweet
## 20 Apple    Red  Large Not sweet

Define a function named `concept_class` that takes a fruit as input and returns its classification (“Ripe” or “Not Ripe”) based on the specified fruit names.

concept_class <- function(fruit) {
  if (color %in% c("Red", "Purple")) {
    return("Ripe")
  } else {
    return("Not Ripe")
  }
}

#Define a function named hypothesis_class that takes color, size, and sweetness as inputs and returns the classification (“Ripe” or “Not Ripe”) based on specific conditions defined within the function.

# Define the hypothesis class
hypothesis_class <- function(color, size, sweetness) {
  if (color %in% c("Red", "Purple") || (size == "Large" && sweetness == "Sweet")) {
    return("Ripe")
  } else {
    return("Not Ripe")
  }
}

Classify each fruit in the data frame

# Classify each fruit in the data frame
# A loop that iterates over each row of the `fruits_df` data frame and extracts the values for each attribute (fruit, color, #size, sweetness) from the current row of the data frame.
for (i in 1:nrow(fruits_df)) {
  fruit <- fruits_df[i, "Fruit"]
  color <- fruits_df[i, "Color"]
  size <- fruits_df[i, "Size"]
  sweetness <- fruits_df[i, "Sweetness"]
  
# Classify using the hypothesis class
#calls the `hypothesis_class` function with the extracted attribute values and assigns the resulting classification to the 
# 'result' variable.
  result <- hypothesis_class(color, size, sweetness)
  
  # Print the instance and the classification
  cat("Fruit:", fruit, "Classification:", result, "\n")
}

## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Not Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe 
## Fruit: Apple Classification: Ripe

# Define a loss function that calculates the loss incurred by the hypothesis class for a given fruit based on its classification
loss_function <- function(true_class, predicted_class) {
  # Define the loss values
  loss_values <- ifelse(true_class != predicted_class, 1, 0)  # 0-1 loss function
  
  # Calculate the total loss
  total_loss <- sum(loss_values)
  return(total_loss)
}

# Calculate the loss for each fruit in the data frame
# Iterate over each row of the `fruits_df` data frame to calculate the loss for each fruit classification
for (i in 1:nrow(fruits_df)) {
  fruit <- fruits_df[i, "Fruit"]
  color <- fruits_df[i, "Color"]
  size <- fruits_df[i, "Size"]
  sweetness <- fruits_df[i, "Sweetness"]
  
  # True classification based on the concept class
  true_class <- concept_class(color)
  
  # Predicted classification based on the hypothesis class
  predicted_class <- hypothesis_class(color, size, sweetness)
  
  # Calculate loss using the loss function
  fruit_loss <- loss_function(true_class, predicted_class)
  
  # Print the instance, true class, predicted class, and loss
  cat("Fruit: ", fruit, " True Class: ", true_class, " Predicted Class: ", predicted_class, " Loss: ", fruit_loss, "\n")
}

## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Ripe  Loss:  1 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Ripe  Loss:  1 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Ripe  Loss:  1 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Not Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0 
## Fruit:  Apple  True Class:  Not Ripe  Predicted Class:  Ripe  Loss:  1 
## Fruit:  Apple  True Class:  Ripe  Predicted Class:  Ripe  Loss:  0

PAC learning

C mcginnis

4/22/2024

Randomly assign charisteristic to the apple color, size and sweet not sweet.

Create the dataframe named ‘fruits_df by combining the ’fruits’, ‘colors’, sizes’, and ’sweetness into a vector column

Define a function named `concept_class` that takes a fruit as input and returns its classification (“Ripe” or “Not Ripe”) based on the specified fruit names.

Classify each fruit in the data frame

PAC learning

C mcginnis

4/22/2024

Randomly assign charisteristic to the apple color, size and sweet not sweet.

Create the dataframe named ‘fruits_df by combining the ’fruits’, ‘colors’, sizes’, and ’sweetness into a vector column

Define a function named concept_class that takes a fruit as input and returns its classification (“Ripe” or “Not Ripe”) based on the specified fruit names.

Classify each fruit in the data frame

Define a function named `concept_class` that takes a fruit as input and returns its classification (“Ripe” or “Not Ripe”) based on the specified fruit names.