add as_sparse_double()

[EmilHvitfeldt]

to close #15

I tried writing a C version of this function. It did better for smaller lengths and worse for longer. so IMO not worth the complexity considering that this isn't going to be a much used function

# A tibble: 10 × 14
   expression     n      min   median `itr/sec` mem_alloc `gc/sec` n_itr  n_gc
   <bch:expr> <dbl> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl> <int> <dbl>
 1 as_sparse…   1e1   9.96µs  10.87µs    78997.        0B     23.7  9997     3
 2 as_sparse…   1e1    3.2µs   3.61µs   261443.        0B     52.3  9998     2
 3 as_sparse…   1e2  10.17µs  11.07µs    86843.      984B     26.1  9997     3
 4 as_sparse…   1e2   3.69µs    4.1µs   234696.      984B     23.5  9999     1
 5 as_sparse…   1e3   11.6µs  12.87µs    74080.    7.99KB     22.2  9997     3
 6 as_sparse…   1e3   7.22µs      8µs   108443.    7.99KB     21.7  9998     2
 7 as_sparse…   1e4  24.93µs  30.18µs    31959.   78.31KB     76.9  9976    24
 8 as_sparse…   1e4  42.56µs  46.25µs    21150.   78.31KB     48.9  9089    21
 9 as_sparse…   1e5 153.26µs 170.27µs     5809.  781.43KB    161.   2089    58
10 as_sparse…   1e5 390.48µs  409.8µs     2420.  781.43KB     68.1  1030    29
# ℹ 5 more variables: total_time <bch:tm>, result <list>, memory <list>,
#   time <list>, gc <list>