Skip to main content
  • EN
  • Home
  • Forums
  • FME Form
  • Data
  • Way of specifying a floating point attribute's width and precision when converting from csv to shape?

I am looking to convert csv files to point shape files. Each attribute in the shape files require a specific format. Character string, date and double format all create as expected. The trouble comes with the attributes that require to be set as floating point. FME always defaults the floating point width/precision to 13.11 which adds additional zeros on to the values when viewing the attributes in the shape files.

In this case the attributes required to be floating point are x, y, z, roll, pitch and yaw of aerial image photo-centres. Ideally the floating point width/precision needs to be 13.3 for XYZ and 9.4 for roll/pitch/yaw.

The image below shows our validation tool used on the shape files output from FME. With expected fields at the top (green highlight) and fields found in the shape file below (red highlight):

The additional zeros are not seen when viewing the shapefile in FME Data Inspector 2019.1, however they are visible when viewing the shape file in other GIS packages such as ArcMap, ArcPro and QGIS.

When creating float attributes in FME, the width precision columns are greyed out and reading through the documentation, unless I am mistaken, it seems that float width and precision default to 13.11 and cannot be edited.

Has anyone any thoughts as to how this issue might be addressed within FME?

In the Shape file writer feature type, you can specify the attributes as type Number, with a specific width and precision.

That will allow for 13.3 and 9.4

Hope this helps

In the Shape file writer feature type, you can specify the attributes as type Number, with a specific width and precision.

That will allow for 13.3 and 9.4

Hope this helps

Thank you for the response. Unfortunately this was tested and number type in FME actually translates as double rather than floating point in the shape file attribution once exported:

I know Arc for example, automatically changes float type attributes to double once a certain number of characters is reached. FME seems to circumvent this by initially allowing float type over a certain character limit. Unfortunately it seems to be constrained to the 13.11 width/precision.

Hi.

 

I have had the same problem, but the writer doesn't allow me to specify the precision or width, they are in grey and are non-editable...

Hi.

 

I have had the same problem, but the writer doesn't allow me to specify the precision or width, they are in grey and are non-editable...

That's because Double numbers are floating-point - the decimal place could come anywhere. You'll want the number data type, which is fixed-point and lets you define the width and precision.

Reply


Community Stats

30,184
Posts
114,569
Replies
37,846
Members
Terms & ConditionsCookie settings
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.