GPU device selection: default when no env variable set

pyfstat/tcw_fstat_map_funcs.py

@@ -129,7 +129,10 @@ def init_transient_fstat_map_features ( ):
         for n, devn in enumerate(devices):
             logging.debug('device {} model: {}, RAM: {}MB'.format(n,devn.name(),devn.total_memory()/(2.**20) ))
 
-        devnum = int(os.environ['CUDA_DEVICE'])
+        if 'CUDA_DEVICE' in os.environ:
+            devnum = int(os.environ['CUDA_DEVICE'])
+        else:
+            devnum = 0
         devn = drv.Device(devnum)
         logging.info('Choosing CUDA device {}, of {} devices present: {}... (Can be changed through environment variable $CUDA_DEVICE.)'.format(devnum,num_gpus,devn.name()))
         logging.debug('Available GPU memory: {}/{} MB free'.format(drv.mem_get_info()[0]/(2.**20),drv.mem_get_info()[1]/(2.**20)))