Fitting multiple parametric distributions to data and simulate best-fitting distribution • misc

Example usage 1

set.seed(123)
n <- 1000
vector <- rweibull(n, 2, 3)  # Replace with your vector

start <- proc.time()[3]
simulate_function <- misc::fit_param_dist(vector)

##  [1] 17.67329861  0.09061650  0.12501307  4.46394854  0.03302592         NaN
##  [7]  0.08655094  0.04391711         Inf  0.04878161  0.90238403  0.02175046
## [1] "Best distribution based on KL divergence: weibull"

end <- proc.time()[3]
print(paste("Time taken:", end - start))

## [1] "Time taken: 0.385"

simulated_data <- simulate_function(n)  # Generate 100 samples from the best-fit distribution

par(mfrow = c(1, 2))
hist(vector, main = "Original Data", xlab = "Value", ylab = "Frequency")
hist(simulated_data, main = "Simulated Data", xlab = "Value", ylab = "Frequency")

Example usage 2

set.seed(123)
n <- 1000
vector <- rnorm(n)  # Replace with your vector

start <- proc.time()[3]
simulate_function <- misc::fit_param_dist(vector)

##  [1] 12.56009975  0.06457125  9.61265383  9.60274199  9.57615179         NaN
##  [7]         Inf  0.01611113         Inf  0.01283163  0.15734994         Inf
## [1] "Best distribution based on KL divergence: normal"

end <- proc.time()[3]
print(paste("Time taken:", end - start))

## [1] "Time taken: 0.207"

simulated_data <- simulate_function(n)  # Generate 100 samples from the best-fit distribution

par(mfrow = c(1, 2))
hist(vector, main = "Original Data", xlab = "Value", ylab = "Frequency")
hist(simulated_data, main = "Simulated Data", xlab = "Value", ylab = "Frequency")

Example usage 3

# Example usage 1
set.seed(123)
n <- 1000
vector <- rlnorm(n)  # Replace with your vector

start <- proc.time()[3]
simulate_function <- misc::fit_param_dist(vector)

##  [1] 10.15776279  0.03482640  0.04704209  0.70874223  0.05123187         NaN
##  [7]  0.01779376  0.25579394         Inf  0.41587181  0.04209187  0.04123623
## [1] "Best distribution based on KL divergence: log-normal"

end <- proc.time()[3]
print(paste("Time taken:", end - start))

## [1] "Time taken: 0.377"

simulated_data <- simulate_function(n)  # Generate 100 samples from the best-fit distribution

par(mfrow = c(1, 2))
hist(vector, main = "Original Data", xlab = "Value", ylab = "Frequency")
hist(simulated_data, main = "Simulated Data", xlab = "Value", ylab = "Frequency")

Example usage 4

set.seed(123)
n <- 1000
vector <- rbeta(n, 2, 3)  # Replace with your vector

start <- proc.time()[3]
simulate_function <- misc::fit_param_dist(vector, verbose=TRUE)

##  [1] 0.003348187 0.076889442 0.317792831 0.192600862 0.037330018         NaN
##  [7] 0.107711307 0.025396848         Inf 0.016032465         Inf 0.011174278
## [1] "Best distribution based on KL divergence: beta"

end <- proc.time()[3]
print(paste("Time taken:", end - start))

## [1] "Time taken: 0.423999999999999"

simulated_data <- simulate_function(n)  # Generate 100 samples from the best-fit distribution

par(mfrow = c(1, 2))
hist(vector, main = "Original Data", xlab = "Value", ylab = "Frequency")
hist(simulated_data, main = "Simulated Data", xlab = "Value", ylab = "Frequency")