mavextra.py: return None if nan passed as new value to lowpass

similar to what other function in here to, don't poison the lowpass result with nans - just ignore the bad value

mavextra.py

@@ -271,6 +271,8 @@ def second_derivative_9(var, key):
 
 def lowpass(var, key, factor):
     '''a simple lowpass filter'''
+    if isnan(var):
+        return None
     global lowpass_data
     if not key in lowpass_data:
         lowpass_data[key] = var