调用函数根据参数汇总数据帧并将其应用于数据帧和nls

Question

我正在使用天气数据和水电费帐单，并尝试估算非线性回归模型。

我想出一个问题。 我调用的用于计算天气统计信息的函数供暖和制冷天数（HDD和CDD）无法应用于数据框，而nls无法使用它。 显然，我缺少有关函数参数的明显信息。

有人可以指出我下面的HDD和CDD功能出了什么问题吗？

这是一个简单的示例问题，其中包含一些生成伪造的天气和帐单数据的代码。

# Generate Fake Weather Data
CZ<-c(1,2)
Date<-c('2001-01-01','2001-01-02','2001-01-03','2001-01-04')
Weather<-expand.grid(CZ,Date)
names(Weather)<-c("CZ","Date")
Weather$AvgTemp<-rnorm(8,mean= 60,sd=20)

#Generate Fake Billing Data
ID<-as.numeric(1:10)
CZ<-c(1,2)
StartDate<-'2001-01-01'
EndDate<-'2001-02-01'
FakeBilling<-data.frame(cbind(ID,CZ,StartDate,EndDate))
FakeBilling$KWH<-rnorm(10,mean=1000, sd=200)

#Heating and cooling degree functions
HDD<- function(b,CZ,StartDate,EndDate) {
    Temps<-Weather$AvgTemp[Weather$CZ==CZ&as.Date(Weather$Date) >=as.Date(StartDate) &     as.Date(Weather$Date) < as.Date(EndDate)];

    sum((b-Temps)/(1+exp(-5*(b-Temps))))
}


CDD <- function(b,CZ,StartDate,EndDate) {
Temps<- Weather$AvgTemp[as.character(Weather$CZ)==as.character(CZ) &     as.Date(Weather$Date) >=as.Date(StartDate)& as.Date(Weather$Date) < as.Date(EndDate)]

    sum((Temps-b)/(1+exp(-5*(Temps-b))))
}

#these work
HDD(60,1,'2001-01-01','2001-02-01')
# [1] 29.34333
CDD(60,1,'2001-01-01','2001-02-01')
# [1] 53.49393

# This does not. Lots of warnings about length
HDD(60,FakeBilling$CZ,FakeBilling$StartDate,FakeBilling$EndDate)
# [1] NA
# Warning messages:
#   1: In is.na(e1) | is.na(e2) :
#   longer object length is not a multiple of shorter object length
# 2: In `==.default`(Weather$CZ, CZ) :
#   longer object length is not a multiple of shorter object length
# 3: In `>=.default`(as.Date(Weather$Date), as.Date(StartDate)) :
#   longer object length is not a multiple of shorter object length
# 4: In `<.default`(as.Date(Weather$Date), as.Date(EndDate)) :
#   longer object length is not a multiple of shorter object length

# Would like to run this but get similar error.
nls(KWH~load + heatload*(HDD(base,CZ,StartDate,EndDate)) ,start=c(load=200,     heatload=.1,base=65), data=FakeBilling, na.action=na.omit)
# Error in numericDeriv(form[[3L]], names(ind), env) : 
#   Missing value or an infinity produced when evaluating the model
# In addition: Warning messages:
#   1: In is.na(e1) | is.na(e2) :
#   longer object length is not a multiple of shorter object length
# 2: In `==.default`(Weather$CZ, CZ) :
#   longer object length is not a multiple of shorter object length
# 3: In `>=.default`(as.Date(Weather$Date), as.Date(StartDate)) :
#   longer object length is not a multiple of shorter object length
# 4: In `<.default`(as.Date(Weather$Date), as.Date(EndDate)) :
#   longer object length is not a multiple of shorter object length
# 5: In is.na(e1) | is.na(e2) :
#   longer object length is not a multiple of shorter object length
# 6: In `==.default`(Weather$CZ, CZ) :
#   longer object length is not a multiple of shorter object length
# 7: In `>=.default`(as.Date(Weather$Date), as.Date(StartDate)) :
#   longer object length is not a multiple of shorter object length
# 8: In `<.default`(as.Date(Weather$Date), as.Date(EndDate)) :
#   longer object length is not a multiple of shorter object length

Answer 1

您的函数未设置为矢量化函数。 您可以使用mapply() 。

with(FakeBilling, mapply(HDD, b = 60, CZ = CZ, StartDate = StartDate, EndDate = EndDate))
#----
[1] 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434

还有一个Vectorize()函数可提供相同的结果：

HDDvec <- Vectorize(HDD)
HDDvec(60,FakeBilling$CZ,FakeBilling$StartDate,FakeBilling$EndDate)
#----
[1] 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434 29.33481 13.39434