R中嵌套For循环的替代方法
我有两个数据集: 竞争对手数据-包含给定产品的竞争对手以及收集竞争对手价格的价格和日期 产品价格-每次价格变动的日期R中嵌套For循环的替代方法,r,for-loop,dplyr,R,For Loop,Dplyr,我有两个数据集: 竞争对手数据-包含给定产品的竞争对手以及收集竞争对手价格的价格和日期 产品价格-每次价格变动的日期 competitor_data <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'), crawl_date=c("2014-0
competitor_data <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
crawl_date=c("2014-04-05", "2014-04-22", "2014-05-05", "2014-05-22","2014-06-05", "2014-06-22",
"2014-05-08", "2014-06-17", "2014-06-09", "2014-06-14","2014-07-01", "2014-08-04"),
competitor =c("amazon","apple","google","facebook","alibaba","tencent","ebay","bestbuy","gamespot","louis vuitton","gucci","tesla"),
competitor_price =c(2.5,2.35,1.99,2.01,2.22,2.52,5.32,5.56,5.01,6.01,5.86,5.96), stringsAsFactors=FALSE)
competitor_data$crawl_date = as.Date(competitor_data$crawl_date)
competitor_data下面的解决方案使用dplyr连接进行匹配。(注意:我将“crawl_date”更改为“date”,这样dplyr join将自动选择匹配的列
by=c('productId'='productId', date'='crawl_date')
作为要联接的参数
competitor_data <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
date=c("2014-04-05", "2014-04-22", "2014-05-05", "2014-05-22","2014-06-05", "2014-06-22",
"2014-05-08", "2014-06-17", "2014-06-09", "2014-06-14","2014-07-01", "2014-08-04"),
competitor =c("amazon","apple","google","facebook","alibaba","tencent","ebay","bestbuy","ga**strong text**mespot","louis vuitton","gucci","tesla"),
competitor_price =c(2.5,2.35,1.99,2.01,2.22,2.52,5.32,5.56,5.01,6.01,5.86,5.96), stringsAsFactors=FALSE)
competitor_data$date = as.Date(competitor_data$date)
product_price <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
date=c("2014-05-05", "2014-06-22", "2014-07-05", "2014-08-31","2014-05-03", "2014-02-22",
"2014-05-21", "2014-06-19", "2014-03-09", "2014-06-22","2014-07-03", "2014-09-08"),
price =c(2.12,2.31,2.29,2.01,2.04,2.09,5.22,5.36,5.21,5.91,5.36,5.56), stringsAsFactors=FALSE)
product_price$date = as.Date(product_price$date)
require(dplyr)
joined <- product_price %>% left_join(competitor_data)
joined$leader <- as.integer(joined$price <= joined$competitor_price)
joined
competitor\u data下面的解决方案使用dplyr join进行匹配。(注意:我将“crawl\u date”更改为“date”,以便dplyr join自动选择匹配的列。它可以通过以下方式显式匹配
by=c('productId'='productId', date'='crawl_date')
作为要联接的参数
competitor_data <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
date=c("2014-04-05", "2014-04-22", "2014-05-05", "2014-05-22","2014-06-05", "2014-06-22",
"2014-05-08", "2014-06-17", "2014-06-09", "2014-06-14","2014-07-01", "2014-08-04"),
competitor =c("amazon","apple","google","facebook","alibaba","tencent","ebay","bestbuy","ga**strong text**mespot","louis vuitton","gucci","tesla"),
competitor_price =c(2.5,2.35,1.99,2.01,2.22,2.52,5.32,5.56,5.01,6.01,5.86,5.96), stringsAsFactors=FALSE)
competitor_data$date = as.Date(competitor_data$date)
product_price <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
date=c("2014-05-05", "2014-06-22", "2014-07-05", "2014-08-31","2014-05-03", "2014-02-22",
"2014-05-21", "2014-06-19", "2014-03-09", "2014-06-22","2014-07-03", "2014-09-08"),
price =c(2.12,2.31,2.29,2.01,2.04,2.09,5.22,5.36,5.21,5.91,5.36,5.56), stringsAsFactors=FALSE)
product_price$date = as.Date(product_price$date)
require(dplyr)
joined <- product_price %>% left_join(competitor_data)
joined$leader <- as.integer(joined$price <= joined$competitor_price)
joined
competitor_datacompetitor_datacompetitor_data为什么不按产品id和日期合并这两个数据集,然后比较这两个价格列,因为爬网日期不一定映射到日期。请查看我的代码中的if语句。因此,您是在下一个最近的日期选择价格,因此在合并后,使用最后一个观察值填写NAS的rward函数这仍然使用for循环,对吗?您可以发布您的解决方案吗?为什么不按产品id和日期合并这两个数据集,然后比较两个价格列,因为爬网日期不一定映射到日期。请查看我的代码中的if语句。因此,您将在下一个最近的日期选择价格,所以在合并之后使用最后一个观察结转函数来填写NAS。这仍然使用for循环,对吗?你能发布你的解决方案吗?缺少的是我的if语句。日期和爬网日期不一定相同。对于给定的日期,我们采用最接近的爬网日期(日期之前的爬网日期)。请查看我的if语句。我输入了逻辑。缺少的是我的if语句。日期和爬网日期不一定相同。对于给定的日期,我们使用最接近的爬网日期(日期之前的爬网日期)。请查看我的if语句。我输入了逻辑。
productId date price competitor competitor_price leader
1 banana 2014-05-05 2.12 google 1.99 0
2 banana 2014-06-22 2.31 tencent 2.52 1
3 banana 2014-07-05 2.29 <NA> NA NA
4 banana 2014-08-31 2.01 <NA> NA NA
5 banana 2014-05-03 2.04 <NA> NA NA
6 banana 2014-02-22 2.09 <NA> NA NA
7 fig 2014-05-21 5.22 <NA> NA NA
8 fig 2014-06-19 5.36 <NA> NA NA
9 fig 2014-03-09 5.21 <NA> NA NA
10 fig 2014-06-22 5.91 <NA> NA NA
11 fig 2014-07-03 5.36 <NA> NA NA
12 fig 2014-09-08 5.56 <NA> NA NA
>
competitor_data <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
crawl_date=c("2014-04-05", "2014-04-22", "2014-05-05", "2014-05-22","2014-06-05", "2014-06-22",
"2014-05-08", "2014-06-17", "2014-06-09", "2014-06-14","2014-07-01", "2014-08-04"),
competitor =c("amazon","apple","google","facebook","alibaba","tencent","ebay","bestbuy","gamespot","louis vuitton","gucci","tesla"),
competitor_price =c(2.5,2.35,1.99,2.01,2.22,2.52,5.32,5.56,5.01,6.01,5.86,5.96), stringsAsFactors=FALSE)
competitor_data$crawl_date = as.Date(competitor_data$crawl_date)
#
product_price <- data.frame(productId=c('banana', 'banana','banana', 'banana','banana', 'banana','fig', 'fig','fig', 'fig','fig', 'fig'),
date=c("2014-05-05", "2014-06-22", "2014-07-05", "2014-08-31","2014-05-03", "2014-02-22",
"2014-05-21", "2014-06-19", "2014-03-09", "2014-06-22","2014-07-03", "2014-09-08"),
price =c(2.12,2.31,2.29,2.01,2.04,2.09,5.22,5.36,5.21,5.91,5.36,5.56), stringsAsFactors=FALSE)
product_price$date = as.Date(product_price$date)
## fill in NAs
f <- function(..., lead = NA) {
# f(NA, 1, NA, 2, NA, NA, lead = NULL)
x <- c(lead, c(...))
head(zoo::na.locf(zoo::na.locf(x, na.rm = FALSE), fromLast = TRUE),
if (is.null(lead)) length(x) else -length(lead))
}
dd <- merge(product_price, competitor_data,
by.y = c('productId', 'crawl_date'),
by.x = c('productId', 'date'), all = TRUE)
dd$competitor_price <-
unlist(sapply(split(dd$competitor_price, dd$productId), f))
dd$price_leader <- +(dd$price <= dd$competitor_price)
(res1 <- `rownames<-`(dd[!is.na(dd$price_leader), -4], NULL))
# productId date price competitor_price price_leader
# 1 banana 2014-02-22 2.09 2.50 1
# 2 banana 2014-05-03 2.04 2.35 1
# 3 banana 2014-05-05 2.12 2.35 1
# 4 banana 2014-06-22 2.31 2.22 0
# 5 banana 2014-07-05 2.29 2.52 1
# 6 banana 2014-08-31 2.01 2.52 1
# 7 fig 2014-03-09 5.21 5.32 1
# 8 fig 2014-05-21 5.22 5.32 1
# 9 fig 2014-06-19 5.36 5.56 1
# 10 fig 2014-06-22 5.91 5.56 0
# 11 fig 2014-07-03 5.36 5.86 1
# 12 fig 2014-09-08 5.56 5.96 1
res0 <- `rownames<-`(all_competitive_data[
order(all_competitive_data$productId, all_competitive_data$date), ], NULL)
all.equal(res0, res1)
# [1] TRUE
library('dplyr')
dd <- full_join(product_price, competitor_data,
by = c(
'productId' = 'productId',
'date' = 'crawl_date'
)
) %>% arrange(productId, date)
dd %>% group_by(productId) %>%
mutate(
competitor_price = f(competitor_price),
price_leader = as.integer(price <= competitor_price)
) %>% filter(!is.na(price_leader)) %>% select(-competitor)
# Source: local data frame [12 x 5]
# Groups: productId [2]
#
# productId date price competitor_price price_leader
# <chr> <date> <dbl> <dbl> <int>
# 1 banana 2014-02-22 2.09 2.50 1
# 2 banana 2014-05-03 2.04 2.35 1
# 3 banana 2014-05-05 2.12 2.35 1
# 4 banana 2014-06-22 2.31 2.22 0
# 5 banana 2014-07-05 2.29 2.52 1
# 6 banana 2014-08-31 2.01 2.52 1
# 7 fig 2014-03-09 5.21 5.32 1
# 8 fig 2014-05-21 5.22 5.32 1
# 9 fig 2014-06-19 5.36 5.56 1
# 10 fig 2014-06-22 5.91 5.56 0
# 11 fig 2014-07-03 5.36 5.86 1
# 12 fig 2014-09-08 5.56 5.96 1