darioappsilon

ggplot(mtcars, aes(x = mpg, y = hp)) + 
  geom_point(color = "#0099f9", size = 5) +
  geom_point(aes(size = qsec, color = cyl)) +
  labs(
    x = "Miles per Gallon",
    y = "Horse power"
  ) +
  theme(
    axis.title.x = element_text(color = "#0099f9", size = 16, face = "bold"),
    axis.title.y = element_text(color = "#0099f9", size = 16, face = "italic")

darioappsilon

ggplot(mtcars, aes(x = mpg, y = hp)) + 
  geom_point(aes(size = qsec, color = cyl)) +
  theme(legend.position = "top")

darioappsilon

ggplot(mtcars, aes(x = mpg, y = hp)) + 
  geom_point(size = 5, color = "#0099f9") + 
  geom_rug()

darioappsilon

library(ggplot2)

# Generate synthetic data with a clear linear relationship
x <- seq(from = 1, to = 300)
y <- rnorm(n = 300, mean = x + 2, sd = 25)

# Convert to dataframe
simple_lr_data <- data.frame(x, y)

# Visualize as scatter plot

darioappsilon

# Calculate coefficients
b1 <- (sum((x - mean(x)) * (y - mean(y)))) / (sum((x - mean(x))^2))
b0 <- mean(y) - b1 * mean(x)

# Define function for generating predictions
simple_lr_predict <- function(x) {
  return(b0 + b1 * x)
}

# Apply simple_lr_predict() to input data

darioappsilon

library(reshape)

# Load in th dataset
df <- read.csv("Fish.csv")

# Remove target variable
temp_df <- subset(df, select = -c(Weight))
melt_df <- melt(temp_df)

# Draw boxplot

darioappsilon

library(caTools)
set.seed(42)

# Train/Test split in 70:30 ratio
sample_split <- sample.split(Y = df$Weight, SplitRatio = 0.7)
train_set <- subset(x = df, sample_split == TRUE)
test_set <- subset(x = df, sample_split == FALSE)

# Fit the model and obtain summary
model <- lm(Weight ~ ., data = train_set)

darioappsilon

# Get residuals
lm_residuals <- as.data.frame(residuals(model))

# Visualize residuals
ggplot(lm_residuals, aes(residuals(model))) +
  geom_histogram(fill = "#0099f9", color = "black") +
  theme_classic() +
  labs(title = "Residuals plot")

darioappsilon

# Make predictions on the test set
predictions <- predict(model, test_set)

# Convert to dataframe
eval <- cbind(test_set$Weight, predictions)
colnames(eval) <- c("Y", "Yhat")
eval <- as.data.frame(eval)
head(eval)

darioappsilon

# Evaluate model
mse <- mean((eval$Y - eval$Yhat)^2)
rmse <- sqrt(mse)