dchall81,008 post(s) |
I'm ecstatic that progress is being made! This is a major step for me to move to M9. Using the Edit Query button to see how the SQL is written, I have expanded the query to create the indexes. I have also added a new field to calculate acreage within the query. Someday I hope to use the query to apply thematic style to the resulting drawing (StyleAreaColorBack), set layer transparency, and create thematic labels. Here is the .prj file that came with the original parcels PROJCS["NAD_1983_StatePlane_Texas_South_Central_FIPS_4204_Feet", GEOGCS["GCS_North_American_1983", DATUM["D_North_American_1983", SPHEROID["GRS_1980",6378137.0,298.257222101]], PRIMEM["Greenwich",0.0], UNIT["Degree",0.0174532925199433]], PROJECTION["Lambert_Conformal_Conic"], PARAMETER["False_Easting",1968500.0], PARAMETER["False_Northing",13123333.33333333], PARAMETER["Central_Meridian",-99.0], PARAMETER["Standard_Parallel_1",28.38333333333333], PARAMETER["Standard_Parallel_2",30.28333333333333], PARAMETER["Latitude_Of_Origin",27.83333333333333], UNIT["Foot_US",0.3048006096012192]] Comparing that to the projection elements previously posted, there are some differences. Is there one correct way to match the Manifold projection to the original prj? False Easting and Northing are considerably different values from the prj to the M9 projection; however, when I open the original and new projected drawing in a map and make one layer translucent, the parcels lie identically on top of each other - at least to the exhaustion of my interest in zooming (873,998:1). Yes the new drawing created by the query is the same one that took 29 minutes to render. This new one, based on the index on PROP_ID in both tables, render in the hundredths of a second. In the interest of science, I am trying to reproduce my 29 minute rendering results... Here's what I got. In rereading my previous posts, here, you are correct that it was a drawing created from a query rather than the new table. The index on Sheet1 was created using the Add Identity button in the Schema and going with defaults. The log does not track how long it takes to plow through the join, so I did not get a time on that. I'd say roughly a half hour for that. Then I opened the drawing created from the query. Render time is listed as 647.961 seconds, but I have a screenshot taken at almost 15 minutes, so something isn't measuring the same. Then I zoomed in to 1:65,000. Render time for that zoom was 410.63 seconds. Then I right-clicked the bottom tab on the drawing and selected zoom again. That render time took 685 seconds. It is almost comical watching the parcels render at such a slow rate. I do try to keep things simple when first trying something. That is why I tried only selecting one field for the original join. I have 42 fields in the Sheet1 db. Are you suggesting I itemize those in the query rather than using SELECT * ? SELECT [Sheet1$].[prop_id],[Sheet1$].[owner_tax_yr],[Sheet1$].[file_as_name],[Sheet1$].[legal_acreage],[Sheet1$].[hood_cd],[Sheet1$].[school],[Sheet1$].[city],[Sheet1$].[county],[Sheet1$].[Last_Appr_Date],[Sheet1$].[legal_desc],[Sheet1$].[tract_or_lot],[Sheet1$].[abs_subdv_cd],[Sheet1$].[land_val],[Sheet1$].[imprv_val],[Sheet1$].[market],[Sheet1$].[assessed_val],[Sheet1$].[block],[Sheet1$].[map_id],[Sheet1$].[geo_id],[Sheet1$].[situs_num],[Sheet1$].[situs_street_prefx],[Sheet1$].[situs_street],[Sheet1$].[situs_street_sufix],[Sheet1$].[situs_city],[Sheet1$].[situs_state],[Sheet1$].[situs_zip],[Sheet1$].[addr_line1],[Sheet1$].[addr_line2],[Sheet1$].[addr_line3],[Sheet1$].[addr_city],[Sheet1$].[addr_state],[Sheet1$].[zip],[Sheet1$].[chg_of_owner_id],[Sheet1$].[deed_book_id],[Sheet1$].[deed_book_page],[Sheet1$].[Deed_Date],[Sheet1$].[deed_num],[Sheet1$].[seq_num],[Sheet1$].[state_cd],[Sheet1$].[ls_table],[Sheet1$].[Group_Codes],[Sheet1$].[entities] I would be more afraid of missing something if I did it like that.
|