You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some way to capture whether a date is fuzzy / uncertain / estimated (clearly need to settle on the language too!). This should probably be a range - the default could just be range(2) for certain / uncertain, but allow users to override it with their own custom range as needed. We should also have reasonable but overrideable defaults for the impact of the fuzziness on comparators. The Python edtf implementation pads by 100% of the uncertainty (e.g. 2022-11~ matches the range for 2022-10-1/2022-12-31 for lower_fuzzy() to upper_fuzzy() ) or 2022-11-01/2022-11-30 (lower_strict() to upper_strict()). This makes sense to me, but we should let users choose the amount of padding. A percentage would be easiest to start, with the “known” date in the middle. In the future could allow more control (the possible range might be skewed).
The text was updated successfully, but these errors were encountered:
@ColeDCrawford I've been thinking that for implementing uncertain dates it would help to have classes of uncertain integers that you can do calculations with, which preserve that uncertainty/fuzziness. I was just looking and found the uncertainties python package, wondering what you think about using this: https://uncertainties.readthedocs.io/en/latest/index.html
initial comments from @ColeDCrawford on slack:
The text was updated successfully, but these errors were encountered: