Create melody dataframe

To extract the melody and time in seconds (needed for animation) from the pyramidi output took only a small bit of wrangling. I used a little modulo operation so that I could pivote the notes into a wide format I could experiment with in gganimate.

pyramidi_raw %>% 
dplyr::mutate(
    status = dplyr::if_else(time == 0, 'on', 'off'),
    ii_note = (i_note + 1) %/% 2,
    note_name = stringr::str_remove(note_name, '\\d+')
) %>% dplyr::select(ii_note, note, note_name, t, status) %>%
tidyr::pivot_wider(names_from = status, values_from = t) 

readRDS(url('https://github.com/softloud/digmus/raw/main/outputs/step-output/melody_df.rds'))

A tibble: 12 × 5

ii_note	note	note_name	on	off
<dbl>	<dbl>	<chr>	<dbl>	<dbl>
1	62	D	0.0	1.2
1	69	A	1.2	2.4
1	65	F	2.4	3.6
2	62	D	3.6	4.8
1	61	C#	4.8	6.0
3	62	D	6.0	6.6
1	64	E	6.6	7.2
2	65	F	7.2	8.7
1	67	G	8.7	9.0
3	65	F	9.0	9.3
2	64	E	9.3	9.6
4	62	D	9.6	10.2