Pymc3 DEMetroplis Initialization

[Hamster_on_wheels]

Would setting the start keyword argument for pm.sample work?

I can set the initial point for NUTS with that method:

Pause and resume training a model that can change during training

Train. Pause. Change model (shift data up by 1) Repeat. Result: [shifted_data] The model is y = mx + c. The parameters are reinitialized during every resumption, which is probably bad. New main function: def main(): # set up our data N = 10 # number of data points sigma = 1. # standard deviation of noise x = np.linspace(0., 9., N) mtrue = 0.4 # true gradient ctrue = 3. # true y-intercept truemodel = my_model([mtrue, ctrue], x) # make data np.rand…