head() and tail() drop the class

[dajmcdon]

library(epiprocess)
#> 
#> Attaching package: 'epiprocess'
#> The following object is masked from 'package:stats':
#> 
#>     filter

head(jhu_csse_daily_subset)
#> # A tibble: 6 × 6
#>   geo_value time_value case_rate_7d_av death_rate_7d_av cases cases_7d_av
#>   <chr>     <date>               <dbl>            <dbl> <dbl>       <dbl>
#> 1 ca        2020-03-01         0.00327         0            6        1.29
#> 2 ca        2020-03-02         0.00435         0            4        1.71
#> 3 ca        2020-03-03         0.00617         0            6        2.43
#> 4 ca        2020-03-04         0.00980         0.000363    11        3.86
#> 5 ca        2020-03-05         0.0134          0.000363    10        5.29
#> 6 ca        2020-03-06         0.0200          0.000363    18        7.86
tail(jhu_csse_daily_subset)
#> # A tibble: 6 × 6
#>   geo_value time_value case_rate_7d_av death_rate_7d_av cases cases_7d_av
#>   <chr>     <date>               <dbl>            <dbl> <dbl>       <dbl>
#> 1 tx        2021-12-26            41.5            0.119 12559      12177.
#> 2 tx        2021-12-27            44.7            0.188 17331      13137 
#> 3 tx        2021-12-28            46.9            0.197 14795      13770.
#> 4 tx        2021-12-29            48.9            0.195 14286      14365.
#> 5 tx        2021-12-30            50.4            0.205 15600      14811.
#> 6 tx        2021-12-31            51.4            0.205 15762      15096.

^{Created on 2022-06-05 by the reprex package (v2.0.1)}

These methods both convert to tibble then dispatch to the data.frame methods. They should either use epiprocess:::reclass() afterword, or, likely better, use [.epi_df.

Possibly see here