r-用于生成新列和新数据帧的用户定义函数

Question

我试图定义一个应用于数据帧disease_df的函数，以实现两件事：

• 生成新列cum_days和week。

• 生成新的数据帧occurence_per_day和occurence_per_week。

我的代码是

# Import data and packages
library(outbreaks)
library(lubridate)
library(dplyr)
disease_df <- rename(measles_hagelloch_1861, date = date_of_prodrome)[, 3, drop = FALSE]
disease_df$date = sort(disease_df$date)
disease_df$disease <- 1
# Define functions to construct tables of number of disease occurence per day and per week respectively
function1 <- function(df)
{
df <- get(df)
# First date
first_date <- min(df$date)
# Last date
last_date <- max(df$date)
# Commulative number of days
df$cum_days <- as.numeric(difftime(df$date, first_date, units = 'days'))
# Week of the date
df$week <- ifelse(df$cum_days / 7 <= 1, 1, ceiling(df$cum_days / 7))
# Construct a table of number of disease occurence per day
occurence_per_day <- aggregate(df$disease, by = list(Category = df$date), FUN = sum)
# Construct a table of number of disease occurence per week
occurence_per_week <- aggregate(df$disease, by = list(Category = df$week), FUN = sum)
occurence_per_day
}
function1(disease_df)

当我运行它时，会出现错误Error in get(df) : invalid first argument。

你能帮我实现目标吗？非常感谢！

Answer 1

这里，我们不需要get(当对象名称以字符串形式传递时是必需的。这里，它是未引用的(

function1 <- function(df) {

# First date
first_date <- min(df[["date"]])
# Last date
last_date <- max(df[["date"]])
# Commulative number of days
df$cum_days <- as.numeric(difftime(df[["date"]], first_date, units = 'days'))
# Week of the date
df$week <- ifelse(df[["cum_days"]] / 7 <= 1, 1, ceiling(df[["cum_days"]] / 7))
# Construct a table of number of disease occurence per day
occurence_per_day <- aggregate(df["disease"], by = list(Category = df[["date"]]), FUN = sum)
# Construct a table of number of disease occurence per week
occurence_per_week <- aggregate(df["disease"], 
by = list(Category = df[["week"]]), FUN = sum)
occurence_per_day
}

-测试

function1(disease_df)
#     Category disease
#1  1861-10-30       1
#2  1861-11-01       1
#3  1861-11-07       2
#4  1861-11-08       1
#5  1861-11-11       2
#6  1861-11-12       1
#7  1861-11-13       1
#8  1861-11-15       2
#9  1861-11-17       1
#10 1861-11-18       1
#11 1861-11-19       1
#12 1861-11-20       4
#13 1861-11-21      14
#14 1861-11-22      12
# ..