The true analysts use R, different from computer geeks using Python!
any()&all()Practice (2024.02.19)- Slicing String with
substr()(2023.06.26) ListPractice in R (2022.11.22)Shiny- 1st Trial (2022.05.04)- Scatter Points in a Circle (2021.08.16)
- Permutations and Combinations (2021.04.05)
- Generating Array and Variables by for Loop (2019.12.06)
- Fibonacci Tornado (2017.05.07)
-
Practice to use
any()all()andduplicated()in R- An in-depth study from Lottery Number Generator (2024.02.15)
-
References
- RDocumentation > any: Are Some Values True?
- Statistics Globe > The all & any R Functions | 4 Example Codes
0. Declare a vector
vec <- c(1:10)
1. Understanding the error
# - code : length(nums) == 6 && !(luckyNum %in% nums) # - error message : 'length = 6' in coercion to 'logical(1)' length(vec) # Output the length of the vector length(vec) == 10 # Check if the length is equal to 10 3 %in% vec # Check if 3 is present in the vector !(3 %in% vec) # Check if 3 is not present in the vector length(vec) == 10 && !(3 %in% vec) # Combine length check and presence check # Explanation: # The error occurred because `luckyNum` was a vector, not a scalar value.
[1] 10 [1] TRUE [1] TRUE [1] FALSE [1] FALSE
2. any() & all()
vec > 5 # Check which elements are greater than 5 any(vec > 5) # Check if any element is greater than 5 all(vec > 5) # Check if all elements are greater than 5
[1] FALSE FALSE FALSE FALSE FALSE TRUE TRUE TRUE TRUE TRUE [1] TRUE [1] FALSE
3. duplicated() with any() & all()
vec2 <- c(vec, 9:12) duplicated(vec2) # Check for duplicated elements in the vector any(duplicated(vec2)) # Check if any element is duplicated all(duplicated(vec2)) # Check if all elements are duplicated
[1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE TRUE TRUE FALSE FALSE [1] TRUE [1] FALSE
-
Dealing with strings in R is a hidden trap
str <- "R도 명색이 프로그래밍 언어인데, 문자열 슬라이싱이 안 될 리가 없잖아. 근데 왜 안 돼? 왜 나 괴롭혀?"
Trial 1 : Do like other general languages
str[1:3] # The entire string is considered as the 1st element of a vector. # The 2nd and 3rd elements of the vector are regarded as empty.
[1] "R도 명색이 프로그래밍 언어인데, 문자열 슬라이싱이 안 될 리가 없잖아. 근데 왜 안 돼? 왜 나 괴롭혀?" [2] NA [3] NA
Trial 2 : Use substr()
substr(str, 1, 19) substr(str, 20, 41) substr(str, 42, 100)
[1] "R도 명색이 프로그래밍 언어인데, " [1] "문자열 슬라이싱이 안 될 리가 없잖아. " [1] "근데 왜 안 돼? 왜 나 괴롭혀?"
Trial 2-1 : For loop with substr()
for (i in 1:nchar(str)) { # not length() cat(substr(str, i, i), seq = " ") }
R 도 명 색 이 프 로 그 래 밍 언 어 인 데 , 문 자 열 슬 라 이 싱 이 안 될 리 가 없 잖 아 . 근 데 왜 안 돼 ? 왜 나 괴 롭 혀 ?
Trial 3 : Use strsplit()
strsplit1 <- strsplit(str, split = "[,] |[.] ", fixed = FALSE) strsplit2 <- strsplit(str, split = "[,.] ", fixed = FALSE) strsplit1 strsplit2 strsplit2[1] strsplit2[[1]] strsplit2[[1]][1] cat(strsplit2[[1]]) cat(strsplit2[[1]][1])
[[1]] [1] "R도 명색이 프로그래밍 언어인데" "문자열 슬라이싱이 안 될 리가 없잖아" "근데 왜 안 돼? 왜 나 괴롭혀?" [[1]] [1] "R도 명색이 프로그래밍 언어인데" "문자열 슬라이싱이 안 될 리가 없잖아" "근데 왜 안 돼? 왜 나 괴롭혀?"
[[1]] [1] "R도 명색이 프로그래밍 언어인데" "문자열 슬라이싱이 안 될 리가 없잖아" "근데 왜 안 돼? 왜 나 괴롭혀?" [1] "R도 명색이 프로그래밍 언어인데" "문자열 슬라이싱이 안 될 리가 없잖아" "근데 왜 안 돼? 왜 나 괴롭혀?" [1] "R도 명색이 프로그래밍 언어인데"
R도 명색이 프로그래밍 언어인데 문자열 슬라이싱이 안 될 리가 없잖아 근데 왜 안 돼? 왜 나 괴롭혀? R도 명색이 프로그래밍 언어인데
-
It's not the linked list that we used to know, but the data strurture that consists of key & value!
0. Set a sample data
Chuhan <- list( "ruler" = "Liu Bei", "general" = c("Guan Yu", "Zhang Fei"), "advisor" = "Zhuge Liang" ) Chuhan
$ruler [1] "Liu Bei" $general [1] "Guan Yu" "Zhang Fei" $advisor [1] "Zhuge Liang"1. Read by key
Chuhan["ruler"] Chuhan[1] # the same with Chuhan["ruler"] Chuhan[[1]] Chuhan[2] Chuhan[2][1] Chuhan[[2]][1] print(Chuhan[[2]][1]) cat(Chuhan[[2]][1])
$ruler [1] "Liu Bei" $ruler [1] "Liu Bei" [1] "Liu Bei" $general [1] "Guan Yu" "Zhang Fei" $general [1] "Guan Yu" "Zhang Fei" [1] "Guan Yu" [1] "Guan Yu" Guan Yu2. Read by value
match("Zhuge Liang", Chuhan) # get the index of the value Chuhan[match("Zhuge Liang", Chuhan)] # the key & value from the index names(Chuhan[match("Zhuge Liang", Chuhan)]) # read only the key
[1] 3 $advisor [1] "Zhuge Liang" [1] "advisor"
-
Hello Shiny
-
Reference ☞ https://shiny.rstudio.com/tutorial/written-tutorial/lesson1/
if (!requireNamespace("shiny")) install.packages("shiny") library("shiny") runExample("01_hello")
-
Scatter points in a circle in various ways
-
Use
plotrix0. Call "plotrix" library (install if not exist)
if(!requireNamespace("plotrix")) install.packages("plotrix") library("plotrix")
1. Monte Carlo method 1
r = 10 n = 30000
rr = runif(n, 0, r) # rr : randomly sampled radius rrad = runif(n, 0, 2 * pi) # rrad : randomly sampled radian x = rr * cos(rrad) # yes, I am a math genius! y = rr * sin(rrad)
windows(width = 7, height = 7) plot(x, y, pch = '.', col = "red", main = "1. Monte Carlo method 1") abline(v = -round(r*1.3):round(r*1.3), h = -r:r, col = "gray") draw.circle(0, 0, r) # not exact drawing, crazy
1.1 Fit the circle on the coordinates
windows(width = 7, height = 7) plot(x, y, pch = '.', col = "red", asp = 1, # modify asp(aspect ratio) option as 1 main = "1.1 Monte Carlo method (with modified asp ratio)") abline(v = -round(r*1.3):round(r*1.3), h = -r:r, col = "gray") draw.circle(0, 0, r)
2. Monte Carlo method 2 (disperse the crowded central population)
x = c(); y = c() cnt = 0
while (cnt < n) # insert points only in the circle { temp = runif(2, -r, r) if (temp[1]^2 + temp[2]^2 < r^2) { x = c(x, temp[1]) y = c(y, temp[2]) cnt = cnt + 1 # I miss ++ operator …… } }
windows(width = 7, height = 7) plot(x, y, pch = '.', col = "red", asp = 1, main = "2. Monte Carlo method 2 (disperse the crowded central pop.)") abline(v = -round(r*1.3):round(r*1.3), h = -r:r, col = "gray") draw.circle(0, 0, r)
3. Points with lattice spacing
x = c(); y = c() area = pi * r^2 interval = sqrt(area / n) num = as.integer(floor(2 * r / interval)) temp = c(-r, -r)
for (i in 1:num) { temp[1] = temp[1] + interval for (j in 1:num) { temp[2] = temp[2] + interval if (temp[1]^2 + temp[2]^2 < r^2) { x = c(x, temp[1]) y = c(y, temp[2]) } } temp[2] = -r }
length(x); length(y)
[1] 29988
[1] 29988windows(width = 7, height = 7) plot(x, y, pch = '.', col = "red", asp = 1, main = "3. Points with lattice spacing") abline(v = -round(r*1.3):round(r*1.3), h = -r:r, col = "gray") draw.circle(0, 0, r)
3.1 Points with lattice spacing including outside the circle
x = c(); y = c(); xyCol = c() temp = c(-r, -r)
for (i in 1:num) { temp[1] = temp[1] + interval for (j in 1:num) { temp[2] = temp[2] + interval x = c(x, temp[1]) y = c(y, temp[2]) if (temp[1]^2 + temp[2]^2 < r^2) xyCol = c(xyCol,"red") else xyCol = c(xyCol,"blue") } temp[2] = -r }
length(x); length(y)
[1] 38025
[1] 38025length(xyCol); length(xyCol[xyCol=="red"]); length(xyCol[xyCol=="blue"])
[1] 38025
[1] 29988
[1] 8037windows(width = 7, height = 7) plot(x, y, pch = '.', col = xyCol, asp = 1, main = "3.1 Points with lattice spacing 2") abline(v = -round(r*1.3):round(r*1.3), h = -r:r, col = "gray") draw.circle(0, 0, r)
-
get permutations and combinations
-
using
gtoolsFactorial
factorial(4) # 4! = 4 * 3 * 2 * 1
[1] 24
Permutation
# loading gtools library if (!requireNamespace("gtools")) { install.packages('gtools') } library(gtools) # for using permutations() and combinations()
# ?permutations # permutations(n, r, v=1:n, set=TRUE, repeats.allowed=FALSE) # n Size of the source vector # r Size of the target vectors # v Source vector. Defaults to 1:n # set Logical flag indicating whether duplicates should be removed from the source vector v. Defaults to TRUE. # repeats.allowed Logical flag indicating whether the constructed vectors may include duplicated values. Defaults to FALSE.
balls <- c("Red", "Yellow", "Blue")
permutations(3, 2, v = balls, repeats.allowed = TRUE) # 3Π2
[1,] "Blue" "Blue"
[2,] "Blue" "Red"
[3,] "Blue" "Yellow"
[4,] "Red" "Blue"
[5,] "Red" "Red"
[6,] "Red" "Yellow"
[7,] "Yellow" "Blue"
[8,] "Yellow" "Red"
[9,] "Yellow" "Yellow"permutations(3, 2, v = balls, repeats.allowed = FALSE) # 3P2 permutations(3, 2, v = balls)
[1,] "Blue" "Red"
[2,] "Blue" "Yellow"
[3,] "Red" "Blue"
[4,] "Red" "Yellow"
[5,] "Yellow" "Blue"
[6,] "Yellow" "Red"Combination
combn(balls, 2)
[1,] "Red" "Red" "Yellow"
[2,] "Yellow" "Blue" "Blue"combinations(3, 2, v = balls, repeats.allowed = TRUE) # 3H2
[1,] "Blue" "Blue"
[2,] "Blue" "Red"
[3,] "Blue" "Yellow"
[4,] "Red" "Red"
[5,] "Red" "Yellow"
[6,] "Yellow" "Yellow"combinations(3, 2, v = balls, repeats.allowed = FALSE) # 3C2
[1,] "Blue" "Red"
[2,] "Blue" "Yellow"
[3,] "Red" "Yellow"Number of cases
prod(4, 2) # 4P2 choose(4, 2) # 4C2
[1] 8
[1] 6
-
answer for a question at chatting room
- R array-related data structure is actually defined as vector, matrix and array about each dimension's array.
- I call it just 'array' by common mathematical notion here, but it is different from R's strict data structure definition.
1. generating array by for loop
mylist = c() for (i in 1:10) { mylist[i] = i } mylist
[1] 1 2 3 4 5 6 7 8 9 10
1.1 generating array more efficiently
mylist2 = c(1:10) mylist2
[1] 1 2 3 4 5 6 7 8 9 10
2. generating variable names
for (i in 1:10) { name <- paste("mylist_", i, sep = "") assign(name, c()) }
2.1 generating variable names with considering sort
for (i in 1:10) { if (i < 10) { name <- paste("mylist_0", i, sep = "") } else { name <- paste("mylist_", i, sep = "") } assign(name, c()) }
2.1.1 code improvement trial of 2.1
name_head_original = "mylist_" for (i in 1:10) { if (i < 10) { name_head = paste(name_head_original, "0", sep = "") } else { name_head = name_head_original } name <- paste(name_head, i, sep = "") assign(name, c()) }
Hmm …… is it too much?
It can be more clearly effective when n is larger, but now seems not yet.
-
Generating Fibonacci Series and Fibonacci Coordinates by looping
1. Generating Fibonacci Series
series <- c(1,1) n <- 1000 ## defining the length of the series for (i in 3:n) { series[i] <- series[i-2] + series[i-1] } head(series)
2. Skimming the Movement of Fibonacci Coordinates
## The series' flow : (0,0), (1,0), (1,1), (-1,1), (-1,-2), (4,-2), …… ## Each Coordinate's movement : ## 0 : x = 0, y = 0 ## 1 : x <- x + 1 ## 2 : y <- y + 1 ## 3 : x <- x - 2 ## 4 : y <- y - 3 ## 5 : x <- x + 5There are 4 types of calculation for coordinates' movement.
It seems possible to be realized by looping.2-1. How sort the types of calculation?
## type 1 : %% 4 = 1 ## type 2 : %% 4 = 2 ## type 3 : %% 4 = 3 ## type 4 : %% 4 = 43. Generating Fibonacci Coordinates by Looping
x <- 0 y <- 0 for (j in 2:n) { if (j %% 2 == 1) { x[j] <- x[j-1] if (j %% 4 == 1) { y[j] <- y[j-1] + series[j-1] ## type 1 } else { y[j] <- y[j-1] - series[j-1] ## type 3 } } else if (j %% 2 == 0) { y[j] <- y[j-1] if (j %% 4 == 2) { x[j] <- x[j-1] + series[j-1] ## type 2 } else { x[j] <- x[j-1] - series[j-1] ## type 4 } } }
3-1. Drawing Plot
windows(width=5, height=5) plot(x[1:12], y[1:12], type="l", main="Fibonacci Tornado") abline(h=0, v=0, col="gray", lty=3)
Bonus. Seeing it's Aproximate to the Golden Ratio
fibonacci.ratio <- c() for (k in 1:n) { fibonacci.ratio[k] = series[k+1]/series[k] }
windows(width=10, height=5) par(mfrow=c(1,2)) plot(fibonacci.ratio[1:12], type="l", main="Aproxmate to the Golden Ratio") abline(h=1.618, col="red", lty=3) plot(log(series[1:12]), type="l", main="Natural Logarithm of Fibonacci Series")
