Derek,
Thanks for the detailed reply - it was very informative. My questions
stemmed from a problem with MI's scaled symbol thematic maps. Because MI
uses symbol fonts for this type of map, the symbols cannot be scaled in true
proportion to the variable being mapped, but instead are scaled in crude
font size steps with a maximum (I think) of 72 points.
A true scaled symbol map produces objects (circles, squares, etc.) whose
radius or area is directly proportional to the value they represent (I think
programs like Geosoft, VM, Surfer, etc. can all do this). As a workaround,
I've written a MapBasic routine to create properly scaled buffers, but was
noticing the area of the buffers wasn't exactly equivalent to 2*pi*r2 where
r is the value I'm trying to represent.
Any thoughts on a better way of doing this? Or the chances of a true scaled
symbol routine being included in future versions of MI.
Thanks again,
_____________________________
Tim Warman
Groundwater Geologist/GIS Guy
Richard C. Slade & Associates LLC
North Hollywood, CA, USA
+1 (818) 506 0418
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Thursday, June 28, 2001 8:01 AM
To: Tim Warman
Cc: Mapinfo-L; [EMAIL PROTECTED]
Subject: Re: MI-L area calculations
Tim,
The problem with circles and ellipses is that they represent unprojected
objects. In order to project the data, there needs to be discrete points
that can be projected. Especially in Lat/Long circles and ellipses, as
well as arcs, rectangles and rounded rectangles, should be used mainly for
cosmetic purposes. In the case of a circle in Lat/Long, the circle itself
lies in a plane tanget to the sphere representing the earth, touching it
only at the center of the circle.
You may notice that in many operations (e.g., Combines, Erases, etc), that
any ellipses, rectangles, rounded rectangles, and arcs are first converted
to Regions (or polylines in the case of arcs). This makes the operations
possible in unprojected Lat/Long coordinate systems, and makes the
mathematics easier in projected coordinate systems. Perhaps we could
improve our handling of Ellipses and other objects, but it is not an often
requested item, especially since the existing workaround, using regions and
polylines, appears to be adequate in most cases.
When calculating Buffers, you can provide the number of nodes that will be
created on points. This resolution is more of a "request" for non-point
objects, as other mathematics comes into play to determine the exact number
of nodes created for the buffered object.
As for CartesianArea, the functionality you state is correct MI Pro
behavior. When you update a column in a table, you are using the table's
coordinate system. The Mapper does not come into play in this calcuation.
And since you can't perform CartesianArea on Lat/Long data (its an
operation that just doesn't make sense), the -1 result is correct. There
are a couple of ways around this issue. One is, as you state, to save the
table in a projected coordinate system. The other is to perform the
operations in MapBasic. If you set the MapBasic coordinate system to State
Plane NAD27, then the object as created in MapBasic will be in that
projected coordinate system. Then, you should be able to perform the
CartesianArea function on that MapBasic object.
Derek Snyder
MapInfo Corporation
Mail List:[EMAIL PROTECTED]
From: "Tim Warman" <[EMAIL PROTECTED]> on 06/27/2001
01:24 PM MST
To: "Mapinfo-L" <[EMAIL PROTECTED]>
cc:
Subject: MI-L area calculations
Listers,
I've got a couple of questions relating to buffers and the way MI
calculates
their area.
1. MI seems to approximate circular buffers with multi-sided polygons. Can
it not just draw a circular buffer?
2. I've got a table of buffers whose native projection/datum is Lat/Long
NAD83. I changed the projection of the map window using
Map>Options>Projection to State Plane NAD27. Then, when I try to update a
column in the buffer table using CartesianArea(obj, "sq ft") it returns a
value of -1 in all instances. Is this a bug or "correct MI behaviour"? Do I
have to Save Copy As to a "cartesian" projection?
TIA,
_____________________________
Tim Warman
Groundwater Geologist/GIS Guy
Richard C. Slade & Associates LLC
North Hollywood, CA, USA
+1 (818) 506 0418
_______________________________________________________________________
List hosting provided by Directions Magazine | www.directionsmag.com |
To unsubscribe, send e-mail to [EMAIL PROTECTED] and
put "unsubscribe MapInfo-L" in the message body.
_______________________________________________________________________
List hosting provided by Directions Magazine | www.directionsmag.com |
To unsubscribe, send e-mail to [EMAIL PROTECTED] and
put "unsubscribe MapInfo-L" in the message body.
