R-dplyr相关函数介绍
install.packages('dplyr') library(dplyr)
> order=read.csv('order.csv',header = T) > head(order) orderid customerid campaignid orderdate city state zipcode 1 1002854 45978 2141 2009/10/13 NEWTON MA 2459 2 1002855 125381 2173 2009/10/13 NEW ROCHELLE NY 10804 3 1002856 103122 2141 2011/6/2 MIAMI FL 33137 4 1002857 130980 2173 2009/10/14 E RUTHERFORD NJ 7073 5 1002886 48553 2141 2010/11/19 BALTIMORE MD 21218 6 1002887 106150 2173 2009/10/15 ROWAYTON CT 6853 paymenttype totalprice numorderlines numunits 1 VI 190.00 3 3 2 VI 10.00 1 1 3 AE 35.22 2 2 4 AE 10.00 1 1 5 VI 10.00 1 1 6 AE 10.00 1 1
1、去重distinct
> order<-rbind(order,order[1,])#合并一行 > nrow(order) [1] 100001 > order<-distinct(order)#去重 > nrow(order) [1] 100000
2、抽样
> nrow(sample_frac(order,0.5))#按照百分之五十抽样 [1] 50000
> nrow(sample_n(order,5))#按照记录数抽样 [1] 5
注意：caret包里的createDataPartition分层抽样
> createDataPartition(order$orderid,p = 0.75,list = F)
3、返回最高记录top_n
> top_n(order,3,totalprice)#数据框、返回最高的3条、按照总价 orderid customerid campaignid orderdate city state zipcode 1 1063586 82645 2141 2012/12/12 NEW YORK NY 10012 2 1054380 88800 2141 2012/11/12 NEW YORK NY 10123 3 1061106 72141 2237 2012/12/12 BIRMINGHAM AL 35203 paymenttype totalprice numorderlines numunits 1 AE 6568.62 14 28 2 AE 6606.00 15 15 3 DB 6780.00 1 400
4、筛选
> head(filter(order,city=='NEWTON'))#等价于order[order$city=='NEWTON',] orderid customerid campaignid orderdate city state zipcode paymenttype 1 1002854 45978 2141 2009/10/13 NEWTON MA 2459 VI 2 1007294 163109 2141 2010/11/29 NEWTON MA 2464 MC 3 1003820 43366 2173 2010/4/9 NEWTON MA 2465 VI 4 1003183 43834 2141 2011/1/12 NEWTON MA 2465 MC 5 1005403 108758 2141 2011/1/17 NEWTON MA 2468 VI 6 1013951 25677 2141 2010/8/16 NEWTON MA 2459 VI totalprice numorderlines numunits 1 190.00 3 3 2 67.50 1 1 3 19.68 1 1 4 67.50 4 4 5 20.25 1 1 6 175.00 1 1
> head(filter(order,city %in% c('NEWTON','MIAMI')))#A%in%B orderid customerid campaignid orderdate city state zipcode paymenttype 1 1002854 45978 2141 2009/10/13 NEWTON MA 2459 VI 2 1002856 103122 2141 2011/6/2 MIAMI FL 33137 AE 3 1003630 42957 2141 2010/11/23 MIAMI FL 33156 AE 4 1003173 123834 2141 2011/1/13 MIAMI FL 33131 AE 5 1004168 112887 2236 2011/1/14 MIAMI FL 33143 VI 6 1005959 125317 2173 2010/11/24 MIAMI FL 33143 AE
补充一点：A%in%B
> c('a','b') %in% c('a','c','b') [1] TRUE TRUE > c('a', 'c', 'b') %in% c('a', 'b') [1] TRUE FALSE TRUE
5、select子集
> head(select(order,city))#也可用来重命名select(order,campid=campaigid) city 1 NEWTON 2 NEW ROCHELLE 3 MIAMI 4 E RUTHERFORD 5 BALTIMORE 6 ROWAYTON
选择以某些字符串开头的列
> head(select(order,starts_with('order'))) orderid orderdate 1 1002854 2009/10/13 2 1002855 2009/10/13 3 1002856 2011/6/2 4 1002857 2009/10/14 5 1002886 2010/11/19 6 1002887 2009/10/15
同样的用法有：ends_with（‘id’），contains（‘date’）
6、arrange排序
> head(arrange(order,orderdate))#按日期升序排列
> head(arrange(order,desc(orderdate)))#按日期降序排列
7、敲黑板：summarize
%>% 将前一步的结果传到下一步summarise中以.来指代上一步结果，实现在R中的数据透视需求
> group_by(order, state) %>% summarise(., n = n(), max = max(totalprice), min = min(totalprice), avg = mean(totalprice)) # A tibble: 81 × 5 state n max min avg <fctr> <int> <dbl> <dbl> <dbl> 1 381 2025.00 0 75.64039 2 AA 7 195.00 10 73.79571 3 AB 54 509.35 0 39.26037 4 AE 20 173.55 10 42.04650 5 AK 54 195.00 0 45.52481 6 AL 269 6780.00 0 103.23814 7 AP 10 204.93 10 52.33800 8 AR 115 1350.00 0 59.08183 9 AZ 1063 2795.00 0 60.68234 10 BC 115 625.00 0 50.39348 # ... with 71 more rows
