[Babase] rain gauge rounding
Niki Learn
nlearn at princeton.edu
Wed Jul 29 14:01:20 EDT 2009
On 07/29/2009 12:25:58 PM, Jeanne Altmann wrote:
> Sorry about not having caught this earlier, but please make it that
> the
> data be in mm as is but that rounding is to nearest single decimal
> digit
> (tenths of a mm).
Ok.
Now for the technicalities, for the record.
Niki and I will have to work out re-doing what Tabby and
I did, I think. Because I recall running the weather
data through some program or another and one of the things
that was done was to round the rain data to the nearest
mm. Either we can re-do/re-enter everything, or do something
to just update the rain data. Niki, what do you think?
Niki writes:
Yes, I saw in the old babase emails something about a file you had sent
Tabby that contained rain data with rounded numbers but I am not finding
that file. The one I found in the folder marked as uploaded to babase had
the rain data with 1-2 decimals (depending on year). That's why it seemed
to me that it must have been uploaded that way and that babase then rounded
the numbers. Can you confirm this one way or the other? Either way I have
the file with the unrounded numbers so if we need to do something to that to
make all the numbers have one decimal place, I can easily supply it. I can
even add through June this year with little effort so we're all up to date.
As far as the computer is concerned, it operates a whole
lot faster if we don't have a decimal point and instead
store the rain information in whole units of 1/10th of
a millimeter. It seems to me that this could be
awkward. I don't think we really care about
performance given the amount of data we're looking
at. Still, it seems worth asking weather or not
anybody cares if the units are integers in tenths
of a millimeter or not. Please comment.
Niki writes:
It definitely sounds like having it in tenths of a mm would be confusing.
The temperature in the same table has one decimal place already so I would
stick with the decimal for rain too. I have had no problems with the speed
of this table when using it.
More information about the Babase
mailing list