You should look in the archives, I think that the definition of the
centroid changed in the latest MI releases, and that it had been adressed
in this list ...
The simple definition of the centroid as the average of the min max values
could lead to a centroid external to the polygon... so MI changed it. ...
at least this is my remembering ... better check the archive.
Eric.
-----Message d'origine-----
De: Mansour Shoari [SMTP:[EMAIL PROTECTED]]
Date: jeudi 10 aout 2000 13:37
A: [EMAIL PROTECTED]
Objet: MI Geosoft's Bug or MapInfo's?
Hello List,
We are a mining exploration company and use MapInfo extensively in
conjunction with major earth science applications (Geosoft, ERMapper,
ModelVision, etc). Late June I reported a bug to Geosoft Corp. with regards
to import of MI-TAB file into their earth science application (Oasis
montaj).
The bug is: when i import MI-TAB files into Oasis montaj, the centroid has
a
considerable (depending where in the world your points are) shift in X and
Y
coordinates.
Geosoft programmers have looked into my bug report and below is their
comment! Could anyone from MI Users/Programmers comment on this?
"Mansour, Late in June you spoke to Erik regarding a region (polygon) in
MapInfo and when you open the TAB in Geosoft our point we report in the
database is off from what is reported in MapInfo. The programmers have now
looked at this and I thought that you might be interested in the
conclusions:
"I don't believe that the Centroid position is a value included in a
MapInfo
TAB file. Neither the old MapInfo Translator API (used up to and including
V4.3) or the current, (and better) MITAB libraries return the Centroid
value; (in fact, the function is defined but not yet implemented in MITAB).
It seems the values displayed in MapInfo are calculated from the current
data, and we do the same. Our values are simply the average of the min and
max X and Y values, calculated in double-precision (8 bytes).
The values are most likely different due to numerical calculation precision
errors; perhaps MapInfo converts point location values internally to scaled
short integers, finds the centroid from the converted values, then
transforms back into real numbers. That the difference between our values
and theirs is on the order of part in 50,000 is evidence for this.
In any case, given the location values we get and plot from the TAB, our
values are exact, and the MapInfo ones are wrong. Unless we obtained
MapInfo's exact internal procedure for scaling and calculating the
centroid,
we have no way of getting "their" values, so this is not a "bug" that we
can
fix."
Thank you in advance,
Mansour Shoari
Monopros Limited
Toronto, Canada
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.146 / Virus Database: 70 - Release Date: 09/05/00
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.146 / Virus Database: 70 - Release Date: 09/05/00
----------------------------------------------------------------------
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
