Error-bounded Lossy Data Compressor (for floating-point/integer datasets)
151
stars
56
forks
source link
fix problems in the situation that dataset contains too much zero result in optimize_intervals_float_2D_opt_MSST19 do not work as we expect. #45
Closed
Borelset closed 5 years ago
if *data_pos = 0 , pred_err = Nan, and the optimize_intervals will be so large.
and now zero value will be skiped, and problem solved.