Hey,

Is it possible that I can do non-equidistant time sampling from a .dfsu and when I save this, the resulting file has equidistant and thereby wrong time stamps?

ds = mikeio.read('my.dfsu')
ds = ds.isel(time=[0,10,20])
ds.to_dfs('my_isel.dfsu')

If this is a known issue, I would suggest a warning when saving or even better optional fill with nan.

If this sounds strange, I would dig a bit further to confirm my observation and provide a better example.

Thanks!