Starting from the main directory of the project (we will call it <maindir>):



- To train the model:

    cd neat/
    python src/nn_evolve.py -g <generations> -f <frequency> -o <outputdir> -c <checkpoint>
    
    
    <generations>:  by default is 10 if not specified
    
    <outputdir>:    set it to <maindir>/neat/, i.e. the current directory: use '-o .'
                    if not specified nothing is stored to files
    
    <frequency>:    how frequently store results (checkpoints and best.pickle) form the training (results are stored avery <frequency> generations)
                    by default it is equal to <generations>, and so results are stored only at the end of the training  
    
    <checkpoint>:   if set initialize the neat algorithm (the population) with the result stored in the <checkpoint> file
                    (which should have been created in a previous run of the algorithm in the <outputdir>)
                    
                    
- To try a model:
    
    1) run the torcs simulator with:
        
        torcs
        
        go to quickrace and in the 'select drivers' configuration, set as driver src_server_2
    
    2) in another terminal:
    
       cd torcs-client
       
       ./start.sh -p 3002 -l -w <model> -o <results_file>
       
       <model>:             pickle file containing the controller for the car
                            use '../best.pickle' to try the best model found by the NEAT algorithm
            
       <results_file>:      path to the file where to store the final results of the race
                            (the information which usuallt are used to compute the fitness function)
                            Usually set to '../model_results/test_results'