1. 向量運算 Vector Computation

👉 R:主要運算符號

1.1 向量數學運算 Mathmatical Opeartions

  ■ 剛開始的時候，R的首要功能就是做向量(vector)運算
  ■ 大多數R的運算符號和功能函數都主要都是對向量做計算
  ■ 數學運算符號和功能函數運算之後會產生數值向量

🌻 Bi-Operant Math Operations act Element-wise

c(1, 2, 3, 4) * c(1, 10, 100, 1000)
[1]    1   20  300 4000

🌻 When the lengths of the vectors are different …

c(100, 200, 300, 400) / 5
[1] 20 40 60 80

The shorter vector are repeated silently

c(100, 200, 300, 400) / c(10, 20)
[1] 10 10 30 20

There is a warning when …

c(10,20,30,40,50,60,70,80) + c(1,2,3)
Warning in c(10, 20, 30, 40, 50, 60, 70, 80) + c(1, 2, 3): longer object length is not a
multiple of shorter object length
[1] 11 22 33 41 52 63 71 82



1.2 向量邏輯(條件)運算 Logical/Conditional OP.

Logical Operations do comparisons and produce logical vectors

  ■ 條件運算符號和功能函數運算之後會產生邏輯向量

They compare numerics, strings, factors, Date, …

c(0.1, 0.2, 0.3, 0.4) > c(0, 1, 2, 3)
[1]  TRUE FALSE FALSE FALSE

shorten vectors are also repeated

c(100, 200, 300, 400) > 250
[1] FALSE FALSE  TRUE  TRUE

❓ what happen if I do …

c(200, 300) > c(100, 200, 300, 400)


🌻 Test for equivalence (==) is different from the assignment operator (=)

c('Amy','Bob','Cindy','Danny') == 'Cindy'
[1] FALSE FALSE  TRUE FALSE

🌻 The Set Comparison Operator : %in% 測試向量元件是否屬與某一個集合

c('Amy','Bob','Cindy','Danny') %in% c('Danny','Cindy')
[1] FALSE FALSE  TRUE  TRUE

The above is the same as …

c('Amy','Bob','Cindy','Danny') %in% c('Cindy','Danny')
[1] FALSE FALSE  TRUE  TRUE
  • sequence in the set vector (right hand side) of %in% is not important 集合元件的次序是不重要的

but different from …

c('Amy','Bob','Cindy','Danny') == c('Danny','Cindy')
[1] FALSE FALSE FALSE FALSE
  • == works element wise. thus, the sequence in the right hand side vector of is important

❓ What happen if I do …

c('Amy','Bob','Cindy','Danny') == c('Danny', 'Cindy')

❓ What if …

c('Amy','Bob','Cindy','Danny') %in% c('Danny', 'Cindy')


1.3 名稱指定 vs. 內容相等 Assignment/Equivalence

🌻 2 Notations of Assignment : = 和 <- 的效果是相同的

Prob = c(0.1, 0.2, 0.3, 0.4)
Value <- c(120, 100, -50, -60)
Prob * Value
[1]  12  20 -15 -24
  • the 2 notations are identical

🌻 Assignment (=) is different from Test for Eq. (==)， 但這兩個運算符號的效果完全不一樣

c(0.1, 0.2, 0.3, 0.4) == (1:4)/10
[1] TRUE TRUE TRUE TRUE

❓ What happen if you do

c(0.1, 0.2, 0.3, 0.4) = (1:4)/10

2. 功能與其參數 Functions & Their Arguments

Most R function take vectors as input (usually the first argument.)

val=c(500,20,75,400)

summary functions produce a single summaries/statistics

sum(val)
[1] 995
mean(val)
[1] 248.8

math functions produce vectors applies to every elements

log(val)
[1] 6.215 2.996 4.317 5.991
sqrt(val)
[1] 22.361  4.472  8.660 20.000

功能選項(arguments)：In addition to the input vector, most R function take extra arguments for options

log(val, base=10)
[1] 2.699 1.301 1.875 2.602

💡 Arguments of Functions：
  ■ To be convenient and flexible, most R functions have many arguments with defaults
  ■ Place cursor on the function name and press F1 to see the online help
  ■ Arguments can be given either by name or by position
  ■ Unnamed-arguments must be in their exact position
  ■ Named arguments can be placed in any order


💡 功能選項(arguments)：
  ■ 為了彈性，多數的功能都有很多個選項
  ■ 為了方便性，多數的功能選項都有預設值
  ■ 將滑鼠的遊標放在功能名稱上，按下F1鍵就可以看到功能的定義
  ■ 在功能定義中，每個選項都有一個名稱
  ■ 假如果你根據功能定義之中各選項的次序來設定選項，就可以不用打選項名稱 (call by position)
  ■ 假如果你指定選項名稱，選項的次序就不必和根據功能定義中的選項次序一樣 (call by name)

help(log)

Default argument

log(1000)
[1] 6.908

Argument by position

log(1000, 10)
[1] 3

Argument by names

log(x=1000, base=10)
[1] 3

Argument by names in reverse order

log(base=10, x=1000)
[1] 3

Quite often we need to cascade several functions, for example

x = 10000
mean(log(sqrt(x), base=10))
[1] 2


🌻 Pipe這一個運算符號可以避免功能呼叫過度巢套，讓程式碼比較符合實際的資料運算程序，比較容易讀懂，也比較不容易寫錯

mean(log(sqrt(x),10))
[1] 2

🌻 The Pipe operator %>% would make it easier to apply a series of functions

sqrt(x) %>% log(10) %>% mean
[1] 2