bind_rows mangles yearmon data

[ConorIA]

This bug is probably on the dplyr side, but I figured I would bring it to your attention anyway. The bind_rows command contained in collapse_hcd chews up the yearmon data in the Date column for monthly data:

> l <- read_hcd(f)
> l
# A tibble: 192 × 12
            Date MaxTemp MinTemp MeanTemp ExtremeHigh ExtremeLow TotalRain TotalSnow TotalPrecip
   <S3: yearmon>   <dbl>   <dbl>    <dbl>       <dbl>      <dbl>     <dbl>     <dbl>       <dbl>
1       Jan 1956      NA      NA       NA          NA         NA        NA        NA          NA
2       Feb 1956      NA      NA       NA          NA         NA        NA        NA          NA
3       Mar 1956      NA      NA       NA          NA         NA        NA        NA          NA
4       Apr 1956      NA      NA       NA          NA         NA      84.6       2.8        87.4
5       May 1956      NA      NA       NA          NA         NA     115.3       0.0       115.3
6       Jun 1956      NA      NA       NA          NA         NA      37.3       0.0        37.3
7       Jul 1956      NA      NA       NA          NA         NA      77.0       0.0        77.0
8       Aug 1956      NA      NA       NA          NA         NA     221.5       0.0       221.5
9       Sep 1956      NA      NA       NA          NA         NA      55.4       0.0        55.4
10      Oct 1956      NA      NA       NA          NA         NA      29.7       0.0        29.7
# ... with 182 more rows, and 3 more variables: LastSnowGrnd <int>, MaxGustDir <int>,
#   MaxGustSpeed <chr>
> l <- bind_rows(l)
> l
# A tibble: 192 × 12
       Date MaxTemp MinTemp MeanTemp ExtremeHigh ExtremeLow TotalRain TotalSnow TotalPrecip
      <dbl>   <dbl>   <dbl>    <dbl>       <dbl>      <dbl>     <dbl>     <dbl>       <dbl>
1  1956.000      NA      NA       NA          NA         NA        NA        NA          NA
2  1956.083      NA      NA       NA          NA         NA        NA        NA          NA
3  1956.167      NA      NA       NA          NA         NA        NA        NA          NA
4  1956.250      NA      NA       NA          NA         NA      84.6       2.8        87.4
5  1956.333      NA      NA       NA          NA         NA     115.3       0.0       115.3
6  1956.417      NA      NA       NA          NA         NA      37.3       0.0        37.3
7  1956.500      NA      NA       NA          NA         NA      77.0       0.0        77.0
8  1956.583      NA      NA       NA          NA         NA     221.5       0.0       221.5
9  1956.667      NA      NA       NA          NA         NA      55.4       0.0        55.4
10 1956.750      NA      NA       NA          NA         NA      29.7       0.0        29.7
# ... with 182 more rows, and 3 more variables: LastSnowGrnd <int>, MaxGustDir <int>,
#   MaxGustSpeed <chr>

We can put it back to normal by re-running l$Date <- as.yearmon(l$Date). Fix recommended in #14.

[gavinsimpson]

Thanks for reporting this @ConorIA. It seems it is a missing "feature" in bind_rows() (I'd call this a bug, but...). I've added a comment to an existing open bug on dplyr tidyverse/dplyr#1829