Pyrevit: Query regarding multiprocessing/threading for executing same functions with different arguments

[kuldeep3618x]

As per the code below, i call in the transaction to write shared parameter values thrice for around 1000 rooms. Each transaction takes about 30 seconds, hence total time- 90 seconds.

• Please let me know, if i can reduce time for the processes.
• Alternatively, could I use multiprocessing or threading to reduce #time by using parallel processing for similar functions with independent arguments. If yes could you give me some insight on the usage.

# _Function to write Shared Parameter values_
def set_parameter_by_name(element, parameterName, value):
    element.LookupParameter(parameterName).Set(value)
# _Function to perform Set Parameter By Name as a Transaction_
def set_parameter_by_name_transaction(sample_elements, sample_param_name, sample_param_values):
    t = Transaction(doc, 'script')
    write_param_pass = []
    write_param_fail = []
    t.Start()
    for e in sample_elements:
        try:
            write_param_pass = [set_parameter_by_name(e,sample_param_name,val) for e,val in zip(sample_elements, sample_param_values)]
        except:
            write_param_fail.append(val)
    t.Commit()  

# _Calling function, for clean set of parameter values_
reset_all_room_occupancies = set_parameter_by_name_transaction(rooms, 'Occupant', list_of_default_values)
# _Calling function, for writing available values_
write_occupant_count_to_rooms = set_parameter_by_name_transaction(rooms_with_valid_occupancy_values, 'Occupant', occupant_count_as_per_code_for_writing_to_param)
# _Calling function, for writing 'NA' for values not calculated_
write_occupant_count_NA_to_rooms_with_no_occupancy_type = set_parameter_by_name_transaction(rooms_with_NA_occupancy_values, 'Occupant', room_NA_occupancy_values)

[eirannejad]

Try running every thing you want to do to all the elements, all in one single transaction

t = Transaction(doc, 'script')
t.Start()
# none of these functions should have a transaction nested inside them
reset_all_room_occupancies()
write_occupant_count_to_rooms()
write_occupant_count_NA_to_rooms_with_no_occupancy_type()
t.Commit()  

On your question about threading, Revit is single-threaded for API changes so only the main thread can make changes.

[kuldeep3618x]

I tried running on a single transaction, but is taking the same amount of time. I think getting Room occupancy, as per selected code, for 1000 rooms in under 90 secs, is in itself an achievement. Thank you.

[eirannejad]

Ya not too bad. Generally the Revit API is the bottleneck. Also try using the .get_Parameter() method instead of .LookupParameter. You might see some performance improvements

[kuldeep3618x]

Will definitely try this out