PySwarms version: 1.3.0
Python version: 3.8.3
Operating System: Ubuntu 18.0.4

Description

im doing a two 5x5 covariance (P estimate and Q process noice) tuning for my Unscented Kalman Filter system. im having an error of

`  208             # fmt: off
    209             self.swarm.current_cost = compute_objective_function(self.swarm, objective_func, pool=pool, **kwargs)
--> 210             self.swarm.pbest_pos, self.swarm.pbest_cost = compute_pbest(self.swarm)
    211             # Set best_cost_yet_found for ftol
    212             best_cost_yet_found = self.swarm.best_cost

~/anaconda3/lib/python3.8/site-packages/pyswarms/backend/operators.py in compute_pbest(swarm)
     67         mask_pos = np.repeat(mask_cost[:, np.newaxis], dimensions, axis=1)
     68         # Apply masks
---> 69         new_pbest_pos = np.where(~mask_pos, swarm.pbest_pos, swarm.position)
     70         new_pbest_cost = np.where(
     71             ~mask_cost, swarm.pbest_cost, swarm.current_cost

<__array_function__ internals> in where(*args, **kwargs)

ValueError: operands could not be broadcast together with shapes (1,10,2) (1,10) (1,10) `

What I Did

I first initialize the parameter with 10 dimension of X, 5 for P and 5 for Q with a testing swarm of 1 for testing : swarm_size = 1 dim = 10 # Dimension of X epsilon = 1.0 options = {'c1': 1.5, 'c2':1.5, 'w':0.5} Next, i create an error function to calculate the RMSE based by the covariance P and Q which changed by X as PSO input particles that i dump to yaml, then to be read by my ROS and C++ script to calculate positions of absolute x and y and predicted x and y position. After that, i calculate the error rmse :

`def error1(X):
    P = np.array([X[0],0,0,0,0,0,X[1],0,0,0,0,0,X[2],0,0,0,0,0,X[3],0,0,0,0,0,X[4]])
    P = P.flatten().tolist()
    Q = np.array([X[5],0,0,0,0,0,X[6],0,0,0,0,0,X[7],0,0,0,0,0,X[8],0,0,0,0,0,X[9]])
    Q = Q.flatten().tolist()
    tes_yaml(P,Q)
    launch()
    hasil_ukf, utm, t_state, t_utm = ambil_bag()
    plot(hasil_ukf,utm)

    # Get the overall rmse
    pos = np.array(err2(hasil_ukf, utm, t_state, t_utm))

    return pos`

later, i create an opt function to reduce the distance between my rmse to the target of 0 as to minimize the rmse `

def opt_func(X):
    n_particles = X.shape[0]  # number of particles
    target = np.array([[0]])
    rmse_min = [(error1(X[i]),target ) for i in range(n_particles)]
    return np.array(rmse_min)
`

lastly, i call the optimizer to optimize the covariance with the dimension of X to minimize the RMSE with 1 iteration for testing if it works or not by writing : %%time

Call an instance of PSO

optimizer = ps.single.GlobalBestPSO(n_particles=swarm_size, dimensions=dim, options=options)

Perform optimization

cost, covariance = optimizer.optimize(opt_func, iters=2)

and thats the error i got, i try to use swarm particle to optimize 10 dimension array for 1 output of rmse. any idea how to fix the error of shape?

ljvmiranda921 / pyswarms

Value operand error when computing best cost for arrays #446

Description

What I Did

Call an instance of PSO

Perform optimization