CODEWITHSUNDEEP
rextractsubset
cljsibyl123
cljsibyl123

Reputation: 5

R programming, extract by using [ ]

stocks <- read.delim("stocks.txt") 
the.tickers <-unique(stocks$ticker)
lows <- c() 
highs <- c()
for(ticker in the.tickers) {   
      look.at <-stocks$ticker == ticker  ## i do not know why the code wrote as this. Any one know? 
      lows <- append(lows,min(stocks$low[look.at],na.rm=TRUE)) ## to find lowest in the 'low' column for each ticker  
      highs <-append(highs,max(stocks$high[look.at],na.rm=TRUE)) 
}  ## to find the highest in the 'high' column for each ticker
the.tickers

My question is, variable look.at is a boolean vector, which contains FALSE. Because there is no character showing as ticker. How can stocks$low return a numerical value even though look.at is all FALSE?

Here is a brief summary of stocks.txt data structure:

str(stocks)
'data.frame':   70061 obs. of  8 variables:
 $ ticker  : Factor w/ 23 levels "AAPL","AMGN",..: 12 12 12 12 12 12 12 12 12 12 ...
 $ industry: Factor w/ 7 levels "Banks","Biotechnology",..: 7 7 7 7 7 7 7 7 7 7 ...
 $ date    : Factor w/ 3086 levels "1-Apr-02","1-Apr-03",..: 1291 1086 985 683 580 478 373 269 2948 2843 ...
 $ open    : num  817 811 805 818 827 ...
 $ high    : num  818 819 813 820 827 ...
 $ low     : num  811 806 801 813 817 ...
 $ close   : num  815 811 808 814 822 ...
 $ volume  : num  1464122 2098176 1838552 3099791 1651111 ...

As you can see, look.at is a boolean object. And it is all False

head(look.at,10)
 [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE

This is the return value:

head(stocks$low[look.at],10)
 [1] 14.94 15.06 15.04 15.12 15.20 15.25 14.95 14.41 14.50 14.58

It does not make sense why Stocks$low[look.at] can return value while look.at is all FALSE.

Variable low is a numeric variable that only contains numbers and NA values. 

head(stocks$low, 10)
 [1] 811.44 806.45 801.47 813.34 817.39 822.31 823.67 831.50 825.05 829.58

sample of data

> head(stocks,10)
   ticker   industry      date   open   high    low  close  volume
1    GOOG Technology 20-Mar-13 816.83 817.51 811.44 814.71 1464122
2    GOOG Technology 19-Mar-13 811.24 819.25 806.45 811.32 2098176
3    GOOG Technology 18-Mar-13 805.00 812.76 801.47 807.79 1838552
4    GOOG Technology 15-Mar-13 818.50 820.30 813.34 814.30 3099791
5    GOOG Technology 14-Mar-13 826.99 826.99 817.39 821.54 1651111
6    GOOG Technology 13-Mar-13 827.90 830.69 822.31 825.31 1641413
7    GOOG Technology 12-Mar-13 830.71 831.89 823.67 827.61 2008979
8    GOOG Technology 11-Mar-13 831.69 839.70 831.50 834.82 1595678
9    GOOG Technology  8-Mar-13 834.50 834.92 825.05 831.52 2912283
10   GOOG Technology  7-Mar-13 834.06 836.62 829.58 832.60 2054238

The.tickers as follows. 

> the.tickers <- unique(stocks$ticker)
> the.tickers
 [1] GOOG AAPL Msft C    KEY  WFC  JPM  SO   DUK  D    HE   EIX  LUV  AMGN GILD CELG BIIB CAT  DE   IMO  MRO  HES  YPF 
Levels: AAPL AMGN BIIB C CAT CELG D DE DUK EIX GILD GOOG HE HES IMO JPM KEY LUV MRO Msft SO WFC YPF

Upvotes: 0

Views: 71

Answers (2)

Gopala
Gopala

Reputation: 10483

If you are trying to get low and high price per ticker, here is how I would do it with the dplyr package:

library(dplyr)
stocks %>% group_by(ticker) %>% summarise(low = min(low, na.rm = TRUE), high = max(high, na.rm = TRUE))

Upvotes: 0

patrickmdnet
patrickmdnet

Reputation: 3392

My question is, variable look.at is a boolean value, how can stocks$low pick up a boolean value, and return a numerical value?

look.at is a boolean vector, not a value. The bracket operator applies each value of the boolean vector to each value of the numeric vector and returns a new vector containing the values of stocks$low that correspond to TRUE values within look.at. Example:

> stockprice <- c(1,2,3)
> look.at <- c(T,F,T)
> stockprice[look.at]
[1] 1 3

Mirosław Zalewski makes the great point in a comment that (unlike what I thought), the bracket operation will work even if look.at is longer or shorter than the stockprice vector:

stockprice <- c(1,2,3,4,5)
look.at <- c(T,F,T,T)
stockprice[look.at]
[1] 1 3 4 5

> stockprice <- c(1,2,3)
> look.at <- c(T,F,T,T)
> stockprice[look.at]
[1]  1  3 NA

Upvotes: 1

Related Questions