I have 30 shapefiles of the same points (300) with one column which is the elevation taken from a DEM (ie 30 different DEMs queried with the same point theme). For analysis, I would like all 30 columns to be in one file (e.g. csv, dbf). tried my luck with featuremerger but with no success.
Thanks, Uwe
Best answer by takashi
The BulkAttributeRenamer and the FeatureMerger can do that.
This is a simplified example, has only 3 input tables.
Assume input shape file names are "name1.shp", "name2.shp",...; each input table has attributes named "ID" and "elevation". 1) Add a SHP Reader to read all shape files with "Single Merged Feature Type" option, expose "fme_feature_type" (holding shp file name) on "Feature Type Properties" dialog box.2) Use the BulkAttributeRenamer to rename "elevation" to value of "fme_feature_type" (i.e. shp file name) with these settings. Rename: Selected Attributes Selected Attributes: elevation Action: Regular Expression Replace Test To Find: .+
(dot and plus symbol. It means one or more any characters) String: fme_feature_type
3) Split feature flow into "name1" features and others with the Tester.
4) Merge all features with the FeatureMerger. Requestor to Supplier Join On: "ID" to "ID" Process Duplicate Suppliers: Yes <-- important!
5) Expose renamed attributes (name1, name2, ...) with the AttributeExposer.
Takashi
Each shape file has a name that should then become the column heading in the new file.
Do columns look like this? Source shape files: file name = name1.shp, name2.shp ... name30.shp | elevation | Expected destination table: name* contains "elevation" from "name*.shp". | name1 | name2 | name3 | ... | name30 |
In the source tables, is there a common attribute which can be used for "Join On" of the FeatureMerger?
Takashi
excactly.
each shape file contains in addition to the elevation attribute the same point ID field to join on which should make it easy as it is a straight join except that in e.g. ArcMap one needs to do this one at a time and as the task is very likely to come up again and again, a neat FME solution would be great.
Many thanks, uwe
The BulkAttributeRenamer and the FeatureMerger can do that.
This is a simplified example, has only 3 input tables.
Assume input shape file names are "name1.shp", "name2.shp",...; each input table has attributes named "ID" and "elevation". 1) Add a SHP Reader to read all shape files with "Single Merged Feature Type" option, expose "fme_feature_type" (holding shp file name) on "Feature Type Properties" dialog box.2) Use the BulkAttributeRenamer to rename "elevation" to value of "fme_feature_type" (i.e. shp file name) with these settings. Rename: Selected Attributes Selected Attributes: elevation Action: Regular Expression Replace Test To Find: .+
(dot and plus symbol. It means one or more any characters) String: fme_feature_type
3) Split feature flow into "name1" features and others with the Tester.
4) Merge all features with the FeatureMerger. Requestor to Supplier Join On: "ID" to "ID" Process Duplicate Suppliers: Yes <-- important!
5) Expose renamed attributes (name1, name2, ...) with the AttributeExposer.
Takashi
As I am still working on FME 2010 (large organisations are a bit slow sometimes) there is no BulkAttributeRenamer or even the AttributeExpressionRenamer that is mentioned by Ken in the link in your message :-(
With the browser at home I can also see the workspace picture which I could not at work. I'll have a go at things the next few days but am sure this will do the trick.
Best regards, Uwe
