Differences in positioning on Map do not match true differences after change projection of a theme.

One example:

Take a Drawing [Vertices] with a number of points, associate to it a projection; I did it with some vertices of national geodetic network.

Duplicate it [Vertices 2] and change the projectionof this new Drawing;

Evaluate differences in XX and YY with the query:

SELECT [V1].[x],[V1].[y],[V1].[N] AS [N1],'[V2].[N] AS [N2],

Abs(Round([V1].[X (I)]-[V2].[X (I)],2)) AS [DX], Abs(Round([V1].[Y (I)]-[V2].[Y (I)],2)) AS [DY]

FROM

[Vertices] AS [V1]

 LEFT JOIN [Vertices 2] AS [V2] 

 ON [V1].[N]=[V2].[N]

ORDER BY [V1].[N]

Differences between these two Drawing will not match diffecences measured in a Map (with one of the Drawings projection).

Why?

Are you saying that coordinate differences as reported from an on-the-fly reprojection in a map differ from those generated by full-blown reprojections the components? That might simply be to do with the different mechanisms by which the values are reported from the different contexts. Also, you are rounding in your query, so are you sure that is not the cause of difference?

I'm not 100% sure on what you are doing here, can you explain more about the Map setup? Can you post a .map file?

cuda shuda wuda

I'll post a map and the xml with the projection parameters I used. Coordinate syetems used for comparisons are cartographic (projected) ones. I'm not comparing Lat Long coordinates against meters.

The rounding applied in the SQL have a negligible impact in this case because differences in a MAP are ~0.005m while differences when comparing internal coordinates are ~100m.

Do this mean that processes used to present data projected on-the-fly differs from the one used to change projection of a Drawing?

Paulo,

When you measure in the map, you have 2 choices: measuring in the map's projection or (with the Ctrl key) measuring over the ellipsoid. I'm sure you're using the first option, so the distances reported are in the units of the map's own projection.

The SQL query is doing something else, something a bit odd at present. It's trying to measure the distance between points in different coordinate systems, without establishing a common coordinate system for the distance to be measured in.

Given different coordinate systems for V1 and V2, the expressions

[V1].[X (I)]-[V2].[X (I)]

[V1].[Y (I)]-[V2].[Y (I)]

aren't meaningful.

You'd need either to project the points in V1 to the coordinate system of V2, then measure the distance in terms of V2; or to do the reverse; or to project all the points to a third coordinate system and measure in that.

(One of these is happening automatically when you measure in the map.)

In this case, V1 and V2 are the same Drawing, one each one with a different projection. I just reproject V2 to another projected snational system. I'm interested in comparing precision obteined after a reprojection of a single theme.

But your suggestion makes sense and I'll try to do it using another data set because I do not have access to the same geodetic vertices with different projections (the old and the new).

I think the X (I) and Y (I) values will differ depending on the projection parameters. A unprojected drawing will have X (I) and Y (I) values in degrees. Local Scale and shifts will also have an influence on these values. So basically you're comparing different things with your query. I wonder if Distance() will give you a better handle on the difference?