General question about the TEMP.gdb and modelling with python

[ChristophEn]

Dear all,

I'm new to programing with arcpy and ArcGIS and have a more general sort of question. Using the ModelBuilder I know that intermediate results within models are often stored in the TEMP.gdb. I started scripting a bit with python and arcpy and wondered if you would really use the TEMP.gdb here as well. At the first glance I can't see any disadvantage of storing intermediate results directly in the original datasets. Of course, sometimes it makes sense to not overwrite useful data but I have the impression that the ModelBuilder generally and always refers to the TEMP.gdb. I can't see any sense in that. Is it just a sort implemented precaution of ArcGIS? Or is it really important? Should I store all my intermediate results in an extra .gdb when building models with python?

I am just interested what you as more experienced users think about that. Appreciate your answer.

Thank you in advance. Regards.

[MatthewLewis]

The "in_memory" location within python or model builder has some limitations which can be found in the help files. It is generally recommended that you use the in_memory database for small amounts of data. Using it to store big datasets will destroy your system performance. Once the process has finished (and closed) the temporary data should be deleted. I use "in_memory" most of the time as its far quicker than writing everything to disk.