Hi there!
How do I handle negative temperatures correctly? My example works for positive values but negative values seem to be treated differently.

In fact it looks like a bug. I believe it is due to the order of precedence of the multiplication given to the unit of measurement that is anticipated to the sign. in fact using the brackets it works properly.

sergio

This works.

Thank you for the different replies!
@Jean Giraud: Not to use units is a no go for me. Mars missions have crashed because of this. And I need the units at a different point in my calculation.

@Sergio and alyles: That looks promising and I indeed think it is a bug related to precedence of multiplications. However, in my real calculation I use the temperature as a variable and thus your workarounds do not apply. The workaround with the Delta breaks the function for positive values. Other ideas?

BTW we talk about the magnus formula for saturated vapour pressure over plain water.




I created a bugzilla issue.

Wrote

Thank you for the different replies!
However, in my real calculation I use the temperature as a variable and thus your workarounds do not apply. The workaround with the Delta breaks the function for positive values. Other ideas?

Try to get negative value of temperature into a bracket. Not doing so cause the issue.
Minus sign before temperature without bracket performed as multiplication of 5°C with -1.
5°C is considered as 278.15K in SMath thus that way of defining variable resulted to -278.15K.
Could this be considered as bug, maybe. Maybe not. SMath first multiply units with number just before them I guess.
Before any of operations units are performed first, it seems this makes the contradiction.

Regards

Thanks overlord. This works for me. It just looks confusing to the inexperienced reader.