Tathagatd96

plt.title("TESTING THE MODEL")

#TRAINING INSTANCE
plt.plot(train_inst[:-1],np.sin(train_inst[:-1]),"bo",markersize=15,alpha=0.5,label="TRAINING INST")

#TARGET TO PREDICT 
plt.plot(train_inst[1:],np.sin(train_inst[1:]),"ko",markersize=8,label="TARGET")

#MODEL PREDCTION
plt.plot(train_inst[1:],y_pred[0,:,0],"r.",markersize=7,label="PREDICTIONS")

Tathagatd96

with tf.Session() as sess:
    sess.run(init)
    
    for iter in range(num_iter):
        x_batch , y_batch = ts_data.next_batch(batch_size,num_time_steps)
        
        sess.run(train,feed_dict={x:x_batch,y:y_batch})
        
        if iter %100==0:
        

Tathagatd96

train_inst=np.linspace(5,5+ts_data.resolution*(num_time_steps+1),num_time_steps+1)

Tathagatd96

#MSE
loss=tf.reduce_mean(tf.square(outputs-y))
optimizer=tf.train.AdamOptimizer(learning_rate=learning_rate)

train=optimizer.minimize(loss)

init=tf.global_variables_initializer()


x_new=np.sin(np.array(train_inst[:-1].reshape(-1,num_time_steps,num_inputs)))

Tathagatd96

#ANY RNN CELL TYPE
cell=tf.contrib.rnn.OutputProjectionWrapper(tf.contrib.rnn.GRUCell(num_units=num_neurons,activation=tf.nn.relu),output_size=num_outputs)

outputs,states=tf.nn.dynamic_rnn(cell,x,dtype=tf.float32)

Tathagatd96

num_inputs=1
num_neurons=100
num_outputs=1
learning_rate=0.001
num_iter=5000
batch_size=1

x=tf.placeholder(tf.float32,[None,num_time_steps,num_inputs])
y=tf.placeholder(tf.float32,[None,num_time_steps,num_outputs])

Tathagatd96

plt.plot(ts_data.x_data,ts_data.y_true)
plt.plot(ts.flatten()[1:],y1.flatten(),"g*")
plt.show()

Tathagatd96

num_time_steps=30

y1,y2,ts= ts_data.next_batch(1,num_time_steps,True)

print(ts.flatten().shape)

plt.plot(ts.flatten()[1:],y1.flatten(),"*")
plt.show()

Tathagatd96

ts_data=TimeSeriesData(250,0,10)
plt.plot(ts_data.x_data,ts_data.y_true)
plt.show()

Tathagatd96

class TimeSeriesData():
    def __init__(self,num_points,xmin,xmax):
        self.xmin=xmin
        self.xmax=xmax
        self.num_points=num_points
        self.resolution=(xmax-xmin)/num_points
        self.x_data=np.linspace(xmin,xmax,num_points)
        self.y_true=np.sin(self.x_data)
        
    def ret_true(self,x_series):