The padr R pacakge vignette describes different package functions to pad dates and times around said dates and times.
I am in situations where I'll be tallying events in data frames (ie dplyr::count()
) and will need to plot occurrences, over a period of say... 1 year. When I count the events in a low volume data frame I'll often get single line item results, like this:
library(tidyverse)
library(lubridate)
library(padr)
df <- tibble(col1 = as.Date("2018-10-01"), col2 = "g", col3 = 5)
#> # A tibble: 1 x 3
#> col1 col2 col3
#> <date> <chr> <dbl>
#> 1 2018-10-01 g 5
To plot this with ggplot, over a period of a year, on a monthly basis, requires a data frame of 12 rows. It basically needs to look like this:
#> # A tibble: 12 x 3
#> col1 col2 col3
#> <date> <chr> <dbl>
#> 1 2018-01-01 NA 0
#> 2 2018-02-01 NA 0
#> 3 2018-03-01 NA 0
#> 4 2018-04-01 NA 0
#> 5 2018-05-01 NA 0
#> 6 2018-06-01 NA 0
#> 7 2018-07-01 NA 0
#> 8 2018-08-01 NA 0
#> 9 2018-09-01 NA 0
#> 10 2018-10-01 g 5
#> 11 2018-11-01 NA 0
#> 12 2018-12-01 NA 0
Perhaps padr()
can do this with some combination of the thicken()
and pad()
functions. My attempts are shown below, neither line 3 nor line 4 construct the data frame shown directly above.
How do I construct that data frame direclty above, utilizing padr()
, lubridate()
, tidyverse()
, data.table()
, base R
, or any way you please? Manual entry of each month shall not be considered either, if that needs to be said. Thank you.
df %>%
thicken("year") %>%
# pad(by = "col1") %>% # line 3
# pad(by = "col1_year") %>% # line 4
print()
library(lubridate)
library(tidyverse)
df <- tibble(col1 = as.Date("2018-10-01"), col2 = "g", col3 = 5)
my_year <- year(df$col1[1])
df2 <- tibble(col1 = seq(ymd(paste0(my_year,'-01-01')),ymd(paste0(my_year,'-12-01')), by = '1 month'))
df3 <- merge(df,df2, by ="col1",all.y=TRUE) %>% mutate(col3 = replace_na(col3,0))
df3