I have an algorithm that generates two sequences x_values and y_values of the same length len(x_values) = len(y_values). However, at each different run of the algorithm the length of these two sequences is slightly higher or lower, len(x_values_run1) != len(x_values_run2).

I want to eventually make a plot with a single x-axis sequence and a single y-axis sequence and for this I need to average across all runs. If the sequences all had the same length, then I could simply do np.vstack(list_x_values).mean(axis=0). However, they have differen lengths.

How can I do this? Perhaps with interpolation.

Minimal Working Example

import numpy as np

def algorithm():
    """Creates 2 list of the same length, different each time."""
    length = np.random.randint(low=18, high=22)
    x_values = np.random.randn(length)
    y_values = 2*x_values + 1
    return x_values, y_values

n_runs = 10
results = [algorithm() for _ in range(n_runs)]
list_x_values = [res[0] for res in results]
list_y_values = [res[1] for res in results]

Important Information

The sequence of x values will typically be on a logarithmic scale.