def process(new_data):
nonlocal prev
new_point = add_points(scale_point(prev, percent),
scale_point(new_data, 1.0 - percent))
prev = new_point
return new_point
def process(new_data):
nonlocal prev_prediction
nonlocal est_velocity
current_velocity = add_points(new_data, invert_point(prev_prediction))
est_velocity = add_points(
scale_point(est_velocity, velocity_smoothing_factor),
scale_point(current_velocity, 1.0 - velocity_smoothing_factor))
predicted_point = add_points(prev_prediction, est_velocity)
smoothed_predicted_point = add_points(
scale_point(predicted_point, prediction_factor),
scale_point(prev_prediction, 1.0 - prediction_factor))
prev_prediction = smoothed_predicted_point
return smoothed_predicted_point
def process(new_data):
nonlocal history
history.append(new_data)
if len(history) > len(weights):
history = history[-len(weights):]
p_sum = Point(0, 0)
for point, weight in zip(reversed(history), weights):
p_sum = add_points(p_sum, scale_point(point, weight))
return p_sum
def generate_line_segment(start_point, end_point, steps):
""" Returns an array of points starting at start point, ending at
end_point, with the defined number of steps. Note that
end point is NOT included, so that consecutive line segments
cane be joined easily."""
points = []
total_delta = add_points(invert_point(start_point), end_point)
for i in range(steps):
percent = i / steps
current_delta = scale_point(total_delta, percent)
current_point = add_points(start_point, current_delta)
points.append(current_point)
return points
def process(new_data):
nonlocal history
history.append(new_data)
if len(history) > len(weights):
history = history[-len(weights):]
p_sum = Point(0, 0)
# Ensure the weights are normalized for the available history
normalized_weights = weights[:len(history)]
normalized_weights = [
w / sum(normalized_weights) for w in normalized_weights
]
for point, weight in zip(reversed(history), normalized_weights):
p_sum = add_points(p_sum, scale_point(point, weight))
return p_sum