데이터 과학
데이터테이블
학습 목표
- dplyr 과 hflights 팩키지 1
데이터 분석 시작
dplyr, hflights 팩키지를 불러와서 head와 summary 명령어로 데이터를 살펴본다.
library(dplyr)
library(hflights)
head(hflights)FALSE Year Month DayofMonth DayOfWeek DepTime ArrTime UniqueCarrier
FALSE 5424 2011 1 1 6 1400 1500 AA
FALSE 5425 2011 1 2 7 1401 1501 AA
FALSE 5426 2011 1 3 1 1352 1502 AA
FALSE 5427 2011 1 4 2 1403 1513 AA
FALSE 5428 2011 1 5 3 1405 1507 AA
FALSE 5429 2011 1 6 4 1359 1503 AA
FALSE FlightNum TailNum ActualElapsedTime AirTime ArrDelay DepDelay Origin
FALSE 5424 428 N576AA 60 40 -10 0 IAH
FALSE 5425 428 N557AA 60 45 -9 1 IAH
FALSE 5426 428 N541AA 70 48 -8 -8 IAH
FALSE 5427 428 N403AA 70 39 3 3 IAH
FALSE 5428 428 N492AA 62 44 -3 5 IAH
FALSE 5429 428 N262AA 64 45 -7 -1 IAH
FALSE Dest Distance TaxiIn TaxiOut Cancelled CancellationCode Diverted
FALSE 5424 DFW 224 7 13 0 0
FALSE 5425 DFW 224 6 9 0 0
FALSE 5426 DFW 224 5 17 0 0
FALSE 5427 DFW 224 9 22 0 0
FALSE 5428 DFW 224 9 9 0 0
FALSE 5429 DFW 224 6 13 0 0
summary(hflights)FALSE Year Month DayofMonth DayOfWeek
FALSE Min. :2011 Min. : 1.000 Min. : 1.00 Min. :1.000
FALSE 1st Qu.:2011 1st Qu.: 4.000 1st Qu.: 8.00 1st Qu.:2.000
FALSE Median :2011 Median : 7.000 Median :16.00 Median :4.000
FALSE Mean :2011 Mean : 6.514 Mean :15.74 Mean :3.948
FALSE 3rd Qu.:2011 3rd Qu.: 9.000 3rd Qu.:23.00 3rd Qu.:6.000
FALSE Max. :2011 Max. :12.000 Max. :31.00 Max. :7.000
FALSE
FALSE DepTime ArrTime UniqueCarrier FlightNum
FALSE Min. : 1 Min. : 1 Length:227496 Min. : 1
FALSE 1st Qu.:1021 1st Qu.:1215 Class :character 1st Qu.: 855
FALSE Median :1416 Median :1617 Mode :character Median :1696
FALSE Mean :1396 Mean :1578 Mean :1962
FALSE 3rd Qu.:1801 3rd Qu.:1953 3rd Qu.:2755
FALSE Max. :2400 Max. :2400 Max. :7290
FALSE NA's :2905 NA's :3066
FALSE TailNum ActualElapsedTime AirTime ArrDelay
FALSE Length:227496 Min. : 34.0 Min. : 11.0 Min. :-70.000
FALSE Class :character 1st Qu.: 77.0 1st Qu.: 58.0 1st Qu.: -8.000
FALSE Mode :character Median :128.0 Median :107.0 Median : 0.000
FALSE Mean :129.3 Mean :108.1 Mean : 7.094
FALSE 3rd Qu.:165.0 3rd Qu.:141.0 3rd Qu.: 11.000
FALSE Max. :575.0 Max. :549.0 Max. :978.000
FALSE NA's :3622 NA's :3622 NA's :3622
FALSE DepDelay Origin Dest Distance
FALSE Min. :-33.000 Length:227496 Length:227496 Min. : 79.0
FALSE 1st Qu.: -3.000 Class :character Class :character 1st Qu.: 376.0
FALSE Median : 0.000 Mode :character Mode :character Median : 809.0
FALSE Mean : 9.445 Mean : 787.8
FALSE 3rd Qu.: 9.000 3rd Qu.:1042.0
FALSE Max. :981.000 Max. :3904.0
FALSE NA's :2905
FALSE TaxiIn TaxiOut Cancelled CancellationCode
FALSE Min. : 1.000 Min. : 1.00 Min. :0.00000 Length:227496
FALSE 1st Qu.: 4.000 1st Qu.: 10.00 1st Qu.:0.00000 Class :character
FALSE Median : 5.000 Median : 14.00 Median :0.00000 Mode :character
FALSE Mean : 6.099 Mean : 15.09 Mean :0.01307
FALSE 3rd Qu.: 7.000 3rd Qu.: 18.00 3rd Qu.:0.00000
FALSE Max. :165.000 Max. :163.00 Max. :1.00000
FALSE NA's :3066 NA's :2947
FALSE Diverted
FALSE Min. :0.000000
FALSE 1st Qu.:0.000000
FALSE Median :0.000000
FALSE Mean :0.002853
FALSE 3rd Qu.:0.000000
FALSE Max. :1.000000
FALSE
데이터프레임을 데이터테이블로 변환
tbl_df 함수를 사용해서 데이터프레임(data.frame)을 데이터테이블(data.table)로 변환한다. 데이터테이블은 내부를 살펴보면 데이터프레임을 기본으로 개발자 관점에서 더욱 쉽고, 빠르고, 직관적으로 데이터를 처리할 수 있는 자료형이다.
hflights <- tbl_df(hflights)
hflightsFALSE Source: local data frame [227,496 x 21]
FALSE
FALSE Year Month DayofMonth DayOfWeek DepTime ArrTime UniqueCarrier
FALSE (int) (int) (int) (int) (int) (int) (chr)
FALSE 1 2011 1 1 6 1400 1500 AA
FALSE 2 2011 1 2 7 1401 1501 AA
FALSE 3 2011 1 3 1 1352 1502 AA
FALSE 4 2011 1 4 2 1403 1513 AA
FALSE 5 2011 1 5 3 1405 1507 AA
FALSE 6 2011 1 6 4 1359 1503 AA
FALSE 7 2011 1 7 5 1359 1509 AA
FALSE 8 2011 1 8 6 1355 1454 AA
FALSE 9 2011 1 9 7 1443 1554 AA
FALSE 10 2011 1 10 1 1443 1553 AA
FALSE .. ... ... ... ... ... ... ...
FALSE Variables not shown: FlightNum (int), TailNum (chr), ActualElapsedTime
FALSE (int), AirTime (int), ArrDelay (int), DepDelay (int), Origin (chr), Dest
FALSE (chr), Distance (int), TaxiIn (int), TaxiOut (int), Cancelled (int),
FALSE CancellationCode (chr), Diverted (int)
carriers <- hflights$UniqueCarrier변수값 재부호화(recoding)
데이터 작업 중에 가장 많이 수행하는 작업 중 하나가 변수내부 값을 다른 값으로 재부호화하는 것이다. 예를 들어, 남자를 Male, 여자를 Female처럼 한국어를 영어로 재부호화하거나, KO를 한국, CN을 중국처럼 코드화된 것을 개발자가 이해하기 새롭게 재부호화하는 것이다. car 라이브러리 recode 함수 등 다양한 재부호화 함수가 존재한다.
데이터프레임의 재부호화하는 방법을 살펴보자. 초간략 코드화된 항공사 코드를 개발자가 이해하기 쉬운 코드로 바꿔보자.
lut <- c("AA" = "American", "AS" = "Alaska", "B6" = "JetBlue", "CO" = "Continental",
"DL" = "Delta", "OO" = "SkyWest", "UA" = "United", "US" = "US_Airways",
"WN" = "Southwest", "EV" = "Atlantic_Southeast", "F9" = "Frontier",
"FL" = "AirTran", "MQ" = "American_Eagle", "XE" = "ExpressJet", "YV" = "Mesa")
hflights$UniqueCarrier <- lut[hflights$UniqueCarrier]
glimpse(hflights)FALSE Observations: 227,496
FALSE Variables: 21
FALSE $ Year (int) 2011, 2011, 2011, 2011, 2011, 2011, 2011, 20...
FALSE $ Month (int) 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,...
FALSE $ DayofMonth (int) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 1...
FALSE $ DayOfWeek (int) 6, 7, 1, 2, 3, 4, 5, 6, 7, 1, 2, 3, 4, 5, 6,...
FALSE $ DepTime (int) 1400, 1401, 1352, 1403, 1405, 1359, 1359, 13...
FALSE $ ArrTime (int) 1500, 1501, 1502, 1513, 1507, 1503, 1509, 14...
FALSE $ UniqueCarrier (chr) "American", "American", "American", "America...
FALSE $ FlightNum (int) 428, 428, 428, 428, 428, 428, 428, 428, 428,...
FALSE $ TailNum (chr) "N576AA", "N557AA", "N541AA", "N403AA", "N49...
FALSE $ ActualElapsedTime (int) 60, 60, 70, 70, 62, 64, 70, 59, 71, 70, 70, ...
FALSE $ AirTime (int) 40, 45, 48, 39, 44, 45, 43, 40, 41, 45, 42, ...
FALSE $ ArrDelay (int) -10, -9, -8, 3, -3, -7, -1, -16, 44, 43, 29,...
FALSE $ DepDelay (int) 0, 1, -8, 3, 5, -1, -1, -5, 43, 43, 29, 19, ...
FALSE $ Origin (chr) "IAH", "IAH", "IAH", "IAH", "IAH", "IAH", "I...
FALSE $ Dest (chr) "DFW", "DFW", "DFW", "DFW", "DFW", "DFW", "D...
FALSE $ Distance (int) 224, 224, 224, 224, 224, 224, 224, 224, 224,...
FALSE $ TaxiIn (int) 7, 6, 5, 9, 9, 6, 12, 7, 8, 6, 8, 4, 6, 5, 6...
FALSE $ TaxiOut (int) 13, 9, 17, 22, 9, 13, 15, 12, 22, 19, 20, 11...
FALSE $ Cancelled (int) 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...
FALSE $ CancellationCode (chr) "", "", "", "", "", "", "", "", "", "", "", ...
FALSE $ Diverted (int) 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...
결과를 일별하는데 glimpse 함수가 요긴하게 사용될 수 있다. 또다른 사례로 운항취소 코드를 개발자가 이애하기 쉬운 코드로 재부호화해보자.
lut = c("A"="carrier", "B"="weather", "C"="FFA", "D"="security","E"="not cancelled")
hflights$CancellationCode <- lut[hflights$CancellationCode]
glimpse(hflights)FALSE Observations: 227,496
FALSE Variables: 21
FALSE $ Year (int) 2011, 2011, 2011, 2011, 2011, 2011, 2011, 20...
FALSE $ Month (int) 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,...
FALSE $ DayofMonth (int) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 1...
FALSE $ DayOfWeek (int) 6, 7, 1, 2, 3, 4, 5, 6, 7, 1, 2, 3, 4, 5, 6,...
FALSE $ DepTime (int) 1400, 1401, 1352, 1403, 1405, 1359, 1359, 13...
FALSE $ ArrTime (int) 1500, 1501, 1502, 1513, 1507, 1503, 1509, 14...
FALSE $ UniqueCarrier (chr) "American", "American", "American", "America...
FALSE $ FlightNum (int) 428, 428, 428, 428, 428, 428, 428, 428, 428,...
FALSE $ TailNum (chr) "N576AA", "N557AA", "N541AA", "N403AA", "N49...
FALSE $ ActualElapsedTime (int) 60, 60, 70, 70, 62, 64, 70, 59, 71, 70, 70, ...
FALSE $ AirTime (int) 40, 45, 48, 39, 44, 45, 43, 40, 41, 45, 42, ...
FALSE $ ArrDelay (int) -10, -9, -8, 3, -3, -7, -1, -16, 44, 43, 29,...
FALSE $ DepDelay (int) 0, 1, -8, 3, 5, -1, -1, -5, 43, 43, 29, 19, ...
FALSE $ Origin (chr) "IAH", "IAH", "IAH", "IAH", "IAH", "IAH", "I...
FALSE $ Dest (chr) "DFW", "DFW", "DFW", "DFW", "DFW", "DFW", "D...
FALSE $ Distance (int) 224, 224, 224, 224, 224, 224, 224, 224, 224,...
FALSE $ TaxiIn (int) 7, 6, 5, 9, 9, 6, 12, 7, 8, 6, 8, 4, 6, 5, 6...
FALSE $ TaxiOut (int) 13, 9, 17, 22, 9, 13, 15, 12, 22, 19, 20, 11...
FALSE $ Cancelled (int) 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...
FALSE $ CancellationCode (chr) NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, ...
FALSE $ Diverted (int) 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...