# **【R Language 自學筆記】**
:::info
- 安裝
- 基本操作
- 參數
- 查詢內建
- 數值運算
- print
- 條件判斷
- while vs for
- 產生序列、複製元素
- random numbers
- for迴圈
- 創建向量
- 離開
- Data Structure: list、factor、df...
- Math
- 字串處理
- 內建資料集
- 創建、讀取檔案: .csv、.txt、.xlsx、.json、RSQLite
- 篩選資料: 篩選 [列,行/欄]、排序、output資料到檔案
- 進階 dplyr
:::
<br/>
### 安裝
- R+RStudio
[R下載](https://cran.r-project.org/)
[RStudio 開發環境下載](https://posit.co/download/rstudio-desktop/)
- 開啟 RStudio
- 新增RScript路徑
先找到路徑，複製
本機(右鍵) - 內容
進階系統設定 - 環境變數
新增加入剛才複製的路徑
- 終端機執行
cd 到檔案位置 - Rscript 執行檔案
- VSCode執行
cd 到檔案位置 - Rscript 執行檔案
>PS 如果之後要用VScode跑，要安裝模組
ctrl+enter執行，會詢問你要不要跳到r終端機，點"確認"
<br/>
### 基本操作
ctrl+enter 執行
ctrl+l: 清空Consloe頁面
alt+shift+k 查詢快捷鍵
[風格](http://adv-r.had.co.nz/Style.html)
其實只要你會Python，能轉換滿快的
整體邏輯和Python、pandas、numpy差不多，例如del(num)，在R是rm(num)；type(num)，在R是class(num)
類似在jupyter notebook 使用，R不需要print(num)，只要num就可印出參數
- 參數
```=
number_1 = 6
number_1
# alt + -
number_2 <- 6
number_2
# 移除
rm(number_2)
```
- 查詢內建
```=
help(iris)
?iris
```
- 數值運算
```=
# +
3+2
# -
3-1
# *
3*2
# /
3/2
# ** or ^
3**3
3^3
# %/%
# 取商數，python是 %
5%%2
# %%
# 取餘數，python是 //
5%/%2
# ^0.5
# 開根號
4^0.5
```
- print
= 與 <- 用法相同
```=
name <- "Catalina"
height <- 1.53
weight <- 46
bmi <- weight / height**2
sprintf("%s's BMI is %s", name, bmi)
sprintf("%s's BMI is %.2f", name, bmi)
```
- 條件判斷
```=
(bmi > 18) | (bmi < 20)
```
```=
if (bmi > 18){
print("要減脂肪")
}
```
```=
if (bmi > 18) | (bmi < 20) {
print("剛剛好")
}
```
```=
if (bmi < 17) {
print("你太瘦了")
} else if (17 < bmi & bmi < 20) {
print("剛剛好")
} else {
print("要減脂肪!")
}
```
```=
# 賦予參數值
num = 10
if ((num%%2) == 0){
result = "是偶數"
}else {
result = "是基數"
}
result
```
```=
# 賦予參數值
if (bmi < 17) {
tag = "你太瘦了"
} else if (17 < bmi & bmi < 20) {
tag = "剛剛好"
} else {
tag = "要減脂肪!"
}
tag
```
- while vs for
```=
# while
n = 1
while (n <= 40) {
if (n %% 20 == 0) {
ans <- "yes"
} else if (n %% 5 == 0) {
ans <- "five"
} else if (n %% 4 == 0) {
ans <- "four"
} else {
ans = n
}
print(ans)
n = n + 1
}
# for
for (n in 1:40) {
if (n %% 20 == 0) {
ans <- "yes"
} else if (n %% 5 == 0) {
ans <- "five"
} else if (n %% 4 == 0) {
ans <- "four"
} else {
ans = n
}
print(ans)
}
```
- while
```=
# 印出1~100
n = 1
while (n <= 100) {
print(n)
n = n + 1
}
```
```=
# 印出1+2+...+100
n = 1
sum = 0
while (n <= 100) {
sum = sum + n
n = n + 1
}
print(sum)
```
```=
# 印出2+4+6...+100
even = 2
sum_even = 0
while (even <= 100 & even%%2 == 0) {
sum_even = sum_even + even
even = even + 2
}
print(sum_even)
```
>for 寫法
```=
sum_100 = 0
for (i in seq(2, to = 100, by = 2)) {
sum_100 = sum_100 + i
}
sum_100
sum(seq(2, to = 100, by = 2))
```
印出特定結果
```=
# 避開4、13 floor
n = 1
n_floors = 20
while (n <= n_floors) {
if(n %in% c(4, 13)) {
n = n + 1
next
} else {
print(sprintf("%s 樓", n))
n = n + 1
}
}
```
>for寫法
```=
for (i in 1:20) {
if(n %in% c(4, 13)) {
next
} else {
print(sprintf("%s 樓", i))
}
}
```
練習: 骰子題
```=
# 要骰幾次，才有五個5?
# 方法1
system.time({
dice = 1:6
dice_rolls = c()
five_count = 0
while (TRUE) {
one_dice = sample(dice, 1)
dice_rolls = c(dice_rolls, one_dice)
if (one_dice[1] == 5) {
five_count = five_count + 1
}
if (five_count == 5) {
break
}
}
print(length(dice_rolls))
print(dice_rolls)
})
# 方法2
system.time({
dice = 1:6
dice_rolls = c()
while (TRUE) {
one_dice = sample(dice, 1)
dice_rolls = c(dice_rolls, one_dice)
if (sum(dice_rolls == 5) == 5) {
break
}
}
print(length(dice_rolls))
print(dice_rolls)
})
```
```=
# 要骰幾次，才有連續五個5?
# 方法一
system.time({
dice = 1:6
dice_rolls = c()
five_count = 0
while (TRUE) {
one_dice = sample(dice, 1)
dice_rolls = c(dice_rolls, one_dice)
if (one_dice[1] == 5) {
five_count = five_count + 1
} else {
five_count = 0
}
if (five_count == 5) {
break
}
}
cat("Method 1:\n")
cat("Length:", length(dice_rolls), "\n")
cat("Last 5 elements:", dice_rolls[(length(dice_rolls) - 4):length(dice_rolls)], "\n")
})
cat("Dice Rolls:", dice_rolls, "\n")
# 方法二
system.time({
dice = 1:6
dice_rolls = c()
while (TRUE) {
one_dice = sample(dice, 1)
dice_rolls = c(dice_rolls, one_dice)
# 檢查是否有連續五個5
if (length(dice_rolls) >= 5 && all(dice_rolls[(length(dice_rolls) - 4):length(dice_rolls)] == 5)) {
break
}
}
cat("Method 2:\n")
cat("Length:", length(dice_rolls), "\n")
cat("Last 5 elements:", dice_rolls[(length(dice_rolls) - 4):length(dice_rolls)], "\n")
})
cat("Dice Rolls:", dice_rolls, "\n")
```
- 產生序列、複製元素
```=
# 類似Python np.arange、np.linspace
seq(from = 1, to = 7, by = 2)
seq(from = 1, to = 7, length.out = 3)
```
```=
# 類似Python np.tile
rep(7, times = 7)
rep("77", times = 7)
rep(TRUE, times = 2)
```
- random numbers
```=
set.seed(100)
random_numbers <- sample(1:1000, size = 100, replace = FALSE)
# odds
for (n in random_numbers) {
if (n %% 2 == 1) {
print(n)
}
}
#odds
random_numbers[random_numbers %% 2 == 1 ]
```
```=
# 到31的倍數就停止
l = c()
while (TRUE) {
l_num = sample(1:1000, size = 1)
l = c(l, l_num)
if (l_num %% 31 == 0) {
break
}
}
length(l)
l
```
加上隨機種子
```=
set.seed(100)
random_numbers <- sample(1:1000, size = 100, replace = FALSE)
# odds
odds = c()
for (n in random_numbers) {
if (n %% 2 == 1) {
odds = c(odds, n)
}
}
odds
length(odds)
# odds
odds = c()
n = 1
while (n <= length(random_numbers)) {
random_num = random_numbers[n]
if (random_num %% 2 == 1) {
odds = c(odds, random_num)
}
n = n + 1
}
odds
length(odds)
```
- for迴圈
```=
for (x in seq(from = 1, to = 7, by = 2)) {
print(x)
}
```
1~100 基數個數
```=
# 1~100 基數個數
count_odd = 0
for (n in 1:100){
if (n%%2 == 1){
count_odd = count_odd + 1
}
}
print(count_odd)
```
>while寫法，要多一個參數，指定起始點
```=
count_odd_w = 0
n = 1
while (n <= 100) {
if (n %% 2 == 1) {
count_odd_w = count_odd_w + 1
}
n = n + 1
}
print(count_odd_w)
```
x~y 基數個數
```=
# x~y 基數個數
x = 12
y = 34
count_odd_r = 0
for (n in x:y) {
if (n %% 2 == 1) {
count_odd_r = count_odd_r + 1
}
}
print(count_odd_r)
```
>while寫法，要多一個參數，指定起始點
```=
x = 12
y = 34
n = x
count_odd_r = 0
while (n <= y) {
if (n %% 2 == 1) {
count_odd_r = count_odd_r + 1
}
n = n + 1
}
print(count_odd_r)
```
因數
```=
# x 有多少因數?
x = 33
count_divisor = 0
for (n in 1:x) {
if (x %% n == 0) {
count_divisor = count_divisor + 1
print(n)
}
}
print(count_divisor)
```
>c() 寫法
```=
x = 33
divisors = c()
for (n in 1:x) {
if (x %% n == 0) {
divisors = c(divisors, n)
}
}
divisors
```
>while寫法，要多一個參數，指定起始點
```=
x = 33
n = 1
count_divisor = 0
while(n <= x){
if(x %% n == 0){
count_divisor = count_divisor + 1
print(n)
}
n = n + 1
}
print(count_divisor)
```
質數
```=
# x 有多少質數?
# 質數:除了1和它本身，没有其他正整数能够整除
x = 33
count_divisor = 0
for (n in 1:x) {
print(sprintf("檢查第 %s 次", n))
if (x %% n == 0) {
count_divisor = count_divisor + 1
print(n)
}
}
if (count_divisor == 2) {
ans = sprintf("%s 是質數", x)
} else {
ans = sprintf("%s 不是質數", x)
}
ans
```
```=
# x 有多少質數?
# 質數:除了1和它本身，没有其他正整数能够整除
divisors = c()
count_divisor = 0
for (n in 1:x) {
print(sprintf("檢查第 %s 次", n))
if (x %% n == 0) {
divisors = c(divisors, n)
count_divisor = 0
}
if (length(divisors) > 2) {
break
}
}
if (length(divisors) == 2) {
ans = sprintf("%s 是質數", x)
} else {
ans = sprintf("%s 不是質數", x)
}
ans
```
>while寫法，要多一個參數，指定起始點
```=
x = 11
n = 1
count_divisor = 0
while(n <= x){
print(sprintf("第 %s 次檢查因數", n))
if(x %% n == 0){
count_divisor = count_divisor + 1
print(n)
}
n = n + 1
}
if (count_divisor == 2) {
ans = sprintf("%s 是質數", x)
} else {
ans = sprintf("%s 不是質數", x)
}
ans
```
>while，改成 n <= x**0.5，扣除掉平方根範圍外的，更有效率
```=
x = 21
n = 1
count_divisor = 0
while(n <= x**0.5){
print(sprintf("第 %s 次檢查因數", n))
if(x %% n == 0){
count_divisor = count_divisor + 1
print(n)
}
if (count_divisor > 2){
break
}
n = n + 1
}
if (count_divisor == 2) {
ans = sprintf("%s 是質數", x)
} else {
ans = sprintf("%s 不是質數", x)
}
ans
```
範圍質數
```=
x = 1
y = 5
primes = c()
for (n in x:y){
divisors = c()
for (m in 1:n){
if (n %% m == 0){
divisors = c(divisors, m)
}
}
if (length(divisors)==2){
primes = c(primes, n)
}
}
length(primes)
primes
```
>包成function
```=
count_prime = function(x, y) {
primes = c()
for (n in x:y) {
divisors = c()
for (m in 1:n) {
if (n %% m == 0) {
divisors = c(divisors, m)
}
}
if (length(divisors) == 2) {
primes = c(primes, n)
}
}
result = list(length=length(primes), primes=primes, divisors=divisors)
return(result)
}
result = count_prime(1, 10)
print(result$length)
print(result[[1]])
print(result$primes)
print(result[[2]])
print(result$divisors)
print(result[[3]])
```
- 創建向量
concat
```=
# 類似Python np.array
name <- c("catalina", "tomas", "biga")
name
```
>長度
```=
name <- c("catalina", "tomas", "biga")
length(name)
```
>索引、切割
```=
name <- c("catalina", "tomas", "biga")
name[1]
name[2]
name[c(1,2)]
```
>新增、減少
```=
name <- c("catalina", "tomas", "biga")
# 新增
name = c(name, "maite")
name
# 減少
name = name[-1]
name
name = name[-2]
name
```
>判斷
```=
age <- c(22, 26, 17, 36)
age > 25
```
>合併使用
```=
name <- c("catalina", "tomas", "biga")
age <- c(22, 26, 17)
name[age > 25]
```
- 離開
```=
q()
```
### Data Structure
- list
可迭代，r中可以存儲 key:value
```=
membership <- list(
"Catalina",
2022,
8.7,
c("data department", "F")
)
class(membership)
```
```=
membership[1]，創建了一個新的包含單一元素的資料結構membership[[1]]，直接返回第一個元素本身
result_1 = membership[1]
print(result_1)
# membership[[1]]
result_2 = membership[[1]]
print(result_2)
```
```=
membership[[4]][2]
```
可跌代
```=
for (m in membership) {
print(m)
}
```
key-value
```=
membership_2 <- list(
name= "Catalina",
year= 2022,
level= 8.7,
info = c("data department", "F")
)
class(membership_2)
```
```=
names(membership_2) # get all keys
membership_2[["name"]]
membership_2$name
membership_2$"info"[2]
```
```=
# 新增k、更新、刪除ey，類似python pandas
membership_2[["age"]] = 26
membership_2
membership_2[["age"]] = 18
membership_2
membership_2[["age"]] = NULL
membership_2
```
>大小寫轉換
```=
mem = c("Catalina", "Biga", "Diego")
split = strsplit(mem, " ")
print(split)
# 大寫
upper_mem = toupper(mem)
print(upper_mem)
# 小寫
lower_mem = tolower(mem)
print(lower_mem)
```
- factor
Factor 將文字向量轉換為整數編碼，通常用於統計分析(統計模型、機器學習算法)
為分類變數的資料結構，文字向量可以轉換為order有序或無序、可以設level
```=
level <- factor(c("gold", "silver", "copper"), ordered = TRUE, levels =c("copper", "silver", "gold"))
level
as.numeric(level)
```
```=
level2 = level[-1]
level2
as.numeric(level2)
```
- Data Frame
Data Frame 是一種二維的資料結構（列&欄），類似於表格，每一列可以是不同的型別，通常用於處理結構化的數據
```=
mem = c("Catalina", "Biga", "Diego")
age = c(26,25,32)
year = c(2020, 2021, 2022)
mem_df = data.frame(title=mem, age, year, stringsAsFactors = FALSE)
mem_df
```
```=
mem_df[["title"]]
mem_df$title
mem_df[["age"]]
mem_df$age
```
```=
mem_df[, 1:2]
```
```=
mem_df[c(1,3), 1:3]
mem_df[c(TRUE, FALSE, TRUE, FALSE), 1:3 ]
```
```=
mem_df[mem_df$age == 25, 1:2]
```
查看表
```=
View(mem_df)
```
str 印出資料框的基本結構
```=
str(mem_df)
```
- Matrix
Matrix 具有相同的數據型別，通常用來進行線性代數運算，例如矩陣相乘和轉置
```=
mat_1 = matrix(1:8, nrow = 2)
mat_1
class(mat_1)
```
```=
mat_1 = matrix(1:6)
mat_1
```
```=
mat_1 = matrix(1:8, nrow = 3)
mat_1
class(mat_1)
```
```=
mat_1 = matrix(1:6, nrow=3)
mat_2 = matrix(5:10, nrow=3)
mat_1
mat_2
mat_1 * mat_2
```
```=
t(mat_1 ) %*% mat_2
```
```=
mat_1 <- matrix(1:6, nrow = 3)
mat_2 <- matrix(5:10, nrow = 2)
mat_1 %*% mat_2
mat_1 = matrix(1:6, nrow=3)
mat_2 = matrix(5:10, nrow=3)
t(mat_1 ) %*% mat_2
```
- Array
Array 可以有多於兩個的維度，用來處理高維矩陣的資料結構
```=
arr_1 = array(1:24, dim = c(2, 3, 4))
arr_1
class(arr_1)
```
<br/>
### Math
```=
# 等同python math.sqrt
sqrt(4)
```
```=
round(2.33)
floor(2.33)
ceiling(2.33)
```
```=
e = exp(2.33)
log(e)
```
```=
log10(10)
log10(10**2)
```
```=
mean(1:5)
mean(c(1,2,3))
```
```=
range(2:10)
max(2:10)
min(2:10)
sum(1:10)
```
```=
name = c("catalina", "biga", "tomas")
unique(name)
toupper(name[[1]])
tolower(name[[1]])
```
```=
grep(pattern = "catalina", name)
```
```=
morning = " Good morning, sir! "
substr(morning, start = 2, stop = 5)
substr(morning, start = 6, stop = nchar(morning))
sub(" .* ", "", morning)
```
```=
trimws(morning, which = "left")
trimws(morning, which = "right")
```
<br/>
### 字串處理
```=
paste(n, ", Good morning, sir!", sep = " :)")
paste0(n, ", Good morning, sir!")
```
```=
for (n in name) {
morning <- paste(n, ", Good morning, sir!", sep = " :)")
print(morning)
}
```
```=
sub(pattern = "catalina", replacement = "cata", name)
```
```=
strsplit(name, split = " ")
```
<br/>
### 內建資料集
類似python seaborn資料集，裡面會有一些像kaggle中的資料集
```=
data()
```
```=
# 查看解說
?euro
```
### 創建、讀取檔案
- csv
創建檔案 .csv
```=
members_csv <- "name,age
catalina,26
biga,25
tomas,31"
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\members_csv.csv"
tryCatch({
writeLines(members_csv, file_path)
print("成功存儲")
}, error = function(e) {
print(paste("存儲失敗:", e))
})
```
讀取檔案 .csv
```=
members_csv_df = read.csv(file_path, header=TRUE, sep=",")
members_csv_df
View(members_csv_df)
```
>ps 用","，excel 資料會自動分開
- txt
創建檔案 .txt
```=
members <- "name;age
catalina;26
biga;25
tomas;31"
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\members.txt"
tryCatch({
writeLines(members, file_path)
print("成功存儲")
}, error = function(e) {
print(paste("存儲失敗:", e))
})
```
讀取檔案 .txt
```=members_df = read.table(file_path, header=TRUE, sep=";")
members_df
View(members_df)
```
- txt 文件檔
創建檔案 .txt
```=
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\news.txt"
news <- "There have been racial barriers, and it has been challenging to be accepted as Japanese. That's what a tearful Carolina Shiino said in impeccable Japanese after she was crowned Miss Japan on Monday. The 26-year-old model, who was born in Ukraine, moved to Japan at the age of five and was raised in Nagoya."
writeLines(news, file_path)
```
讀取檔案 .txt
```=
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\news.txt"
news_char <- readLines(file_path)
class(news_char)
length(news_char)
news_char
```
- json 關聯資料
創建檔案 .json
```=
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\members_json.json"
members_json <- '[{"name": "Catalina", "age": 26},
{"name": "Tomas", "age": 32},
{"name": "Biga", "age": 25}]'
tryCatch({
writeLines(members_json, file_path)
print("成功存儲")
}, error = function(e) {
print(paste("存儲失敗:", e))
})
```
讀取檔案 .json
```=
# 先安裝套件
install.packages("jsonlite")
```
```=
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\members_json.json"
library("jsonlite")
members_json_df = fromJSON(file_path)
members_json_df
```
- json 非關聯資料
創建檔案 .json
```=
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\emp_catalina.json"
emp_catalina = '
{
"name": "Catalina",
"age": 26,
"gender": "F",
"department":["data","business"]
}
'
writeLines(emp_catalina, file_path)
```
讀取檔案 .json
```=
# 先安裝套件
install.packages("jsonlite")
```
```=
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\emp_catalina.json"
library(jsonlite)
emp_catalina = fromJSON(file_path)
emp_catalina
class(emp_catalina)
length(emp_catalina)
names(emp_catalina)
```
- 試算表
讀取檔案 試算表
```=
# 先安裝套件
install.packages("readxl")
```
```=
library("readxl")
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\members_xlsx.xlsx"
members_xlsx_df = read_excel(file_path)
members_xlsx_df
```
- RSQLite
創建 RSQLite
```=
# 先安裝套件
install.packages("RSQLite")
```
創建csv
```=
students_csv <- "name, age
catalina, 26
biga, 25
tomas, 32"
file_path <- "C:\\Users\\catal\\OneDrive\\桌面\\students_csv.csv"
tryCatch({
writeLines(students_csv, file_path)
print("成功存儲")
}, error = function(e) {
print(paste("存儲失敗:", e))
})
```
讀取csv
```=
####
students_csv_df = read.csv(file_path, header=TRUE, sep=",")
students_csv_df
```
寫進db
```=
library(DBI)
conn = dbConnect(RSQLite::SQLite(), ":memory:")
dbListTables(conn)
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\students_csv.csv"
students_csv_df = read.csv(file_path, sep = ",", header=TRUE)
dbWriteTable(conn, "students_csv_df", students_csv_df)
dbListTables(conn)
```
讀取、查詢
```
# 讀取
students_csv_df_db = dbReadTable(conn, "students_csv_df")
students_csv_df_db
# dbSendQuery 查詢
sql_query <- "SELECT * FROM students_csv_df WHERE name = 'catalina';"
res <- dbSendQuery(conn, sql_query)
dbFetch(res)
```
```=
# 清理，釋放相關的資源
dbClearResult(res)
# ： 關閉與數據庫的連接，防止留下不必要的打開連接
dbDisconnect(conn)
```
<br/>
### 篩選資料
- 篩選 [列,行/欄]
```=
str(mem_df)
mem_df[, "age"]
mem_df[, c("title","age")]
mem_df$age
```
```=
mem_df
mem_df[mem_df$title == "Catalina",]
mem_df[mem_df$title == "Diego", "age"]
mem_df[2, ]
mem_df[2, c("title","age")]
```
```=
mem_df[mem_df$age %in% c(22,23,24,25), c("title","age")]
mem_df[mem_df$age %in% 22:25, c("title","age")]
mem_df$age %in% 22:25
```
- 排序
按照年齡
```=
mem_df_indices = order(mem_df[, "age"])
mem_df[mem_df_indices, ]
mem_df[order(mem_df[, "age"]), ]
```
```=
selected_rows = members_csv_df[members_csv_df["age"] > 25, ]
members_more_than25_df = data.frame(name = selected_rows["name"], age = selected_rows["age"])
View(members_more_than25_df)
```
- 新增欄位
```=
mem_df$No. = mem_df$age * 10
head(mem_df)
mem_df$dep =""
head(mem_df)
```
- output資料到檔案
.txt
```=
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\members_more_than25_df.txt"
write.table(members_more_than25_df, file = file_path, row.names = FALSE)
```
.csv
```=
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\members_more_than25_df.csv"
write.table(members_more_than25_df, file = file_path, sep = ",", row.names = FALSE)
```
非關聯 .json
```=
emp_catalina = '
{
"name": "Catalina",
"age": 26,
"gender": "F",
"department":["data","business"]
}
'
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\emp_catalina_2.json"
library(jsonlite)
emp_catalina = toJSON(emp_catalina)
writeLines(emp_catalina, file_path)
```
關聯 .json
```=
file_path = "C:\\Users\\catal\\OneDrive\\桌面\\iris_json "
iris_json = toJSON(iris)
writeLines(iris_json, file_path)
```
<br/>
### 進階 dplyr
%>% 符號是magrittr套件（也被稱為"pipe"操作符號）的一部分，讓代碼更易讀和連貫，它允許前一步的結果傳遞給下一步的函數
%>% 快捷鍵: Ctrl+Shift+M
```=
# 先安裝套件
install.packages("dplyr")
```
```=
library("dplyr")
# 原本寫法
select(mem_df, title)
filter(select(mem_df, c("title","age")), age > 25)
# %>% 寫法
mem_df %>%
select(title,age) %>%
filter(age > 25)
```
```=
mem_df %>%
arrange(`age`) # arrange(desc(`age`))
```
```=
mem_df %>%
mutate(No.2 = age * 20)
mem_df %>%
mutate(No.2 = age * 20) %>%
summarise(avg_no.2 = mean(No.2))
```
Sign in with Wallet
Connect another wallet