Sampling validation dataset

Hello,

I would like to have a sampled learning dataset and a sampled validation dataset.

What I did so far is :

I run iota2 with a configuration file IOTA2_first_run.cfg.

I obtained in the folder /dataAppVal, the sampled dataset for the learning and the dataset for the validation. Now, I would like to sample the validation dataset.

In order to do that, I need to create a shapefile that contains the validation dataset. So I need to merge all the /dataAppVal/*_seed_0_val.sqlite and convert the merged_val.sqlite to a merged_val.shp. However I should have correct field names in order to run iota2 again. Indeed, I will run again iota2 in order to sample the validation dataset (split_ground_truth=False). In my configuration file, the groundtruth will now be my validation shapefile. IOTA2_second_run.cfg

My first idea was to use ogrmerge to merge all the sqlite together and them use ogr2ogr to convert to a shapefile. However I don't have the correct field names.

I have a look in the documentation and more precisely on iota2.VectorTools (http://osr-cesbio.ups-tlse.fr/echangeswww/TheiaOSO/iota2_documentation/master/_autosummary/iota2.VectorTools.html)

Should I use iota2.VectorTools.vector_functions.mergeSQLite instead of ogrmerge ? If I give an example of the correct shapefile, is it possible to change the field names ?

Thanks for your help,

Valentine