# ungrouped data statistics computation
data_mean <- function(arr) {
num <- 0
len <- length(arr)
for (i in arr) {
num <- num + i
}
return (num / len)
}
data_median <- function(arr) {
arr_2 <- sort(arr)
len <- length(arr)
if (len %% 2 == 1) {
get <- len %/% 2
return (arr_2[get + 1])
} else {
mid1 <- len / 2
mid2 <- mid1 + 1
return ((arr_2[mid1] + arr_2[mid2]) / 2)
}
}
data_mode <- function(arr) {
freq <- table(arr)
mode_val <- [Link](names(freq)[freq == max(freq)])
return (mode_val)
}
data_variance <- function(arr) {
sum_val <- 0
freq <- 0
len <- length(arr)
d_mean <- data_mean(arr)
for (i in arr) {
sum_val <- sum_val + (i - d_mean)^2
}
return (sum_val / len)
}
data_stddev <- function(arr) {
return (sqrt(data_variance(arr)))
}
data_nthtile <- function(arr, nth, type) {
arr_2 <- sort(arr)
val <- (nth / type) * (length(arr) + 1)
inter <- val - floor(val)
n1 <- arr_2[floor(val)]
n2 <- arr_2[floor(val) + 1]
return (n1 + (inter * (n2 - n1)))
}
arr <- c(92, 92, 89, 88, 81, 91, 93, 90, 95, 87)
data_mean(arr) # or just do mean(arr)
data_median(arr) # might as well be median(arr)
data_mode(arr) # mode here returns its data type, so it is hardcoded here
data_variance(arr) # var(arr) also exists here
data_stddev(arr) # sd(arr) is even here as well
data_nthtile(arr, 3, 4) # quantiles(arr, args) are also built-in here
data_nthtile(arr, 2, 10) # but works differently for when the
data_nthtile(arr, 25, 100) # required is n-tile
# grouped data statistics computation
group_mean <- function(arr) {
sum_val <- 0
freq <- 0
for (i in 1:nrow(arr)) {
temp <- (arr[i, 1] + arr[i, 2]) / 2
sum_val <- sum_val + (arr[i, 3] * temp)
freq <- freq + arr[i, 3]
}
return (sum_val / freq)
}
group_median <- function(arr) {
cf <- cumsum(arr[3])
n <- sum(arr[3])
mi_ind <- which(cf >= n / 2)[1]
L <- arr[mi_ind, 2]
freq <- ifelse(mi_ind == 1, 0, cf[mi_ind - 1])
f_med <- arr[mi_ind, 3]
h <- arr[2, 1] - arr[1, 1] + 1
return (L + ((((n/2) - freq) / f_med) * h))
}
group_mode <- function(arr) {
# get index of the modal class or its freq
mod_ind <- [Link](arr[3])
l_modal <- arr[mod_ind, 2]
f_m <- arr[mod_ind, 3]
f1 <- ifelse(mod_ind == 1, 0, arr[mod_ind - 1, 3])
f2 <- ifelse(mod_ind == nrow(arr), 0, arr[mod_ind + 1, 3])
h <- arr[2, 1] - arr[1, 1] + 1
return (l_modal + ((f_m - f1) / (f_m - f1) + (f_m - f2)) * h)
}
group_variance <- function(arr) {
sum_val <- 0
freq <- 0
g_mu <- group_mean(arr)
for (i in 1:nrow(arr)) {
temp <- (arr[i, 1] + arr[i, 2]) / 2
sum_val <- sum_val + (arr[i, 3] * (temp - g_mu)^2)
freq <- freq + arr[i, 3]
}
return (sum_val / freq)
}
group_stddev <- function(arr) {
return (sqrt(group_variance(arr)))
}
group_nthtile <- function(arr, nth, type) {
freq <- sum(arr[3])
pos <- (nth * freq) / type
cf <- cumsum(arr[3])
index <- which(cf >= pos)[1]
l_qnt <- arr[index, 2]
F_qnt <- ifelse(index == 1, 0, cf[index - 1])
f_qnt <- arr[index, 3]
h <- arr[2, 1] - arr[1, 1] + 1
return (l_qnt + ((pos - F_qnt) / f_qnt) * h)
}
class_data <- [Link] (
up_class = c(20, 25, 30, 35, 40, 45),
lw_class = c(16, 21, 26, 31, 36, 41),
freq = c(2, 7, 14, 8, 8, 1)
)
group_mean(class_data)
group_median(class_data)
group_mode(class_data)
group_variance(class_data)
group_stddev(class_data)
group_nthtile(class_data, 3, 4)
group_nthtile(class_data, 2, 10)
group_nthtile(class_data, 25, 100)
Locsin, Paul Aljon B.
Vallejos, Allein Dion A.
Briones, Dwight Kirby
Quilon, Jommel Sean B.
Lanting, Neil Gabriel S. (Absent)
