Question: Running a learning for additional points after some amount of points was computed

[serhiy-yevtushenko]

Hi

My question is the following:

how one could run a learner again after sampling a certain amount of points.

I have tried following:

run LearnerND for sampling 2000 points.

Saved them to the dataframe

Started new run.

Created new instance of LeanerND with same initial parameters and function.

Loaded points generated from old session using

learner.load_dataframe(initial_df,
                           with_default_function_args=False,
                           point_names=("x", "y"),
                           value_name="value"

)

and tried running

    points = len(initial_df)+110
    runner=BlockingRunner(learner, npoints_goal=points)
    df = learner.to_dataframe(point_names=("x", "y"))

However, I get the exception:

raise ValueError("Point already in triangulation.")

from add_point (line 629, in add_point, called from line 605, in _try_adding_pending_point_to_simplex)

[akhmerov]

Seems like a bug; this should work. Please try to make a self-contained reproducible example.

[serhiy-yevtushenko]

I managed to do reproducible example (it uses fictional function, but real inputs for convex hull boundaries and saved points.

Here it is:

import numpy as np
import pandas as pd
from adaptive import LearnerND, BlockingRunner
from scipy.spatial import ConvexHull

if __name__ == "__main__":

    def some_f(xy):
        x, y = xy
        a = 0.2
        return x + np.exp(-((x**2 + y**2 - 0.75**2) ** 2) / a**4)

    boundaries_list = [
        [0.0, 0.0],
        [100.0, 1.115142836310296e-06],
        [200.0, 2.230285672620592e-06],
        [300.0, 3.3454285089308886e-06],
        [400.0, 4.460571345241184e-06],
        [500.0, 5.57571418155148e-06],
        [600.0, 6.690857017861777e-06],
        [700.0, 7.805999854172073e-06],
        [800.0, 8.921142690482368e-06],
        [900.0, 1.0036285526792665e-05],
        [1000.0, 1.115142836310296e-05],
        [1100.0, 1.2266571199413256e-05],
        [1200.0, 1.3381714035723556e-05],
        [1300.0, 1.449685687203385e-05],
        [1400.0, 1.5611999708344146e-05],
        [1500.0, 1.6727142544654443e-05],
        [1600.0, 1.7842285380964736e-05],
        [1700.0, 1.8957428217275036e-05],
        [1800.0, 2.007257105358533e-05],
        [1900.0, 2.1187713889895622e-05],
        [2000.0, 2.230285672620592e-05],
        [2100.0, 2.3417999562516216e-05],
        [2200.0, 2.4533142398826512e-05],
        [2300.0, 2.564828523513681e-05],
        [2400.0, 2.676342807144711e-05],
        [2500.0, 2.78785709077574e-05],
        [2600.0, 2.89937137440677e-05],
        [2700.0, 3.0108856580378e-05],
        [2800.0, 3.122399941668829e-05],
        [2900.0, 3.233914225299859e-05],
        [3000.0, 3.3454285089308885e-05],
        [3100.0, 3.456942792561918e-05],
        [3200.0, 3.568457076192947e-05],
        [3300.0, 3.679971359823977e-05],
        [3400.0, 3.791485643455007e-05],
        [3500.0, 3.902999927086036e-05],
        [3600.0, 4.014514210717066e-05],
        [3700.0, 4.126028494348096e-05],
        [3800.0, 4.237542777979125e-05],
        [3900.0, 4.349057061610155e-05],
        [4000.0, 4.460571345241184e-05],
        [4100.0, 4.5720856288722135e-05],
        [4200.0, 6.814896913926772e-05],
        [4300.0, 9.86347750047832e-05],
        [4400.0, 0.0001291205808702],
        [4500.0, 0.0001596063867358],
        [4600.0, 0.0001900921926013],
        [4700.0, 0.0002205779984668],
        [4800.0, 0.0002510638043323],
        [4900.0, 0.0002815496101978],
        [5000.0, 0.0003120354160633],
        [5100.0, 0.0003425212219289],
        [5200.0, 0.0003730070277944],
        [5300.0, 0.0004034928336599],
        [5400.0, 0.0004339786395254],
        [5500.0, 0.0004644644453909],
        [5600.0, 0.0004949502512564],
        [5700.0, 0.0005254360571219],
        [5800.0, 0.0005559218629875],
        [5900.0, 0.0007147750527983],
        [5958.286720010206, 0.0191313558591815],
        [5907.6017302122045, 100.0],
        [5900.0, 114.76582808254696],
        [5855.557077950937, 200.0],
        [5802.120964299216, 300.0],
        [5800.0, 303.9180944285528],
        [5747.2470662133965, 400.0],
        [5700.0, 484.0110564366577],
        [5690.889136812792, 500.0],
        [5632.988323616697, 600.0],
        [5600.0, 655.7812882689732],
        [5573.489289877545, 700.0],
        [5512.330765927291, 800.0],
        [5500.0, 819.8310791739173],
        [5449.453695550926, 900.0],
        [5400.0, 976.7273642697368],
        [5384.783431673397, 1000.0],
        [5318.25038673062, 1100.0],
        [5300.0, 1126.9294170460528],
        [5249.768962312114, 1200.0],
        [5200.0, 1270.8848142140498],
        [5179.25599484322, 1300.0],
        [5106.615292424962, 1400.0],
        [5100.0, 1408.9603456835755],
        [5031.748474209412, 1500.0],
        [5000.0, 1541.4981221350852],
        [4954.544472334084, 1600.0],
        [4900.0, 1668.8144779856843],
        [4874.883500782354, 1700.0],
        [4800.0, 1791.1726541541148],
        [4792.629568799029, 1800.0],
        [4707.642393422194, 1900.0],
        [4700.0, 1908.82851256576],
        [4619.75309080672, 2000.0],
        [4600.0, 2022.015570709802],
        [4528.789455590954, 2100.0],
        [4500.0, 2130.934258178381],
        [4434.548819963085, 2200.0],
        [4400.0, 2235.7715308963816],
        [4336.810072561228, 2300.0],
        [4300.0, 2336.709148049797],
        [4235.316140760245, 2400.0],
        [4200.0, 2433.90362418784],
        [4129.769838029961, 2500.0],
        [4100.0, 2527.497213251217],
        [4019.862059018615, 2600.0],
        [4000.0, 2617.630617968991],
        [3905.202836411695, 2700.0],
        [3900.0, 2704.432159381163],
        [3800.0, 2788.0167593762926],
        [3785.344965272928, 2800.0],
        [3700.0, 2868.489829680557],
        [3659.776081181031, 2900.0],
        [3600.0, 2945.946859180247],
        [3527.87276759619, 3000.0],
        [3500.0, 3020.485282516302],
        [3400.0, 3092.192101198906],
        [3388.871711491372, 3100.0],
        [3300.0, 3161.145135978021],
        [3241.856209624924, 3200.0],
        [3200.0, 3227.4215041078746],
        [3100.0, 3291.0867911450027],
        [3085.6746305394568, 3300.0],
        [3000.0, 3352.2106073917794],
        [2918.8280381467725, 3400.0],
        [2900.0, 3410.8539297941925],
        [2800.0, 3467.073798524489],
        [2739.3690425123423, 3500.0],
        [2700.0, 3520.922400948519],
        [2600.0, 3572.446137385158],
        [2544.616895822353, 3600.0],
        [2500.0, 3621.69541538601],
        [2400.0, 3668.7142274073726],
        [2330.724210396878, 3700.0],
        [2300.0, 3713.54356901084],
        [2200.0, 3756.2207047353977],
        [2100.0, 3796.777361608687],
        [2091.822547917115, 3800.0],
        [2000.0, 3835.25044716744],
        [1900.0, 3871.6715389583783],
        [1818.0750272525088, 3900.0],
        [1800.0, 3906.069632377301],
        [1700.0, 3938.470724359623],
        [1600.0, 3968.8952109580896],
        [1500.0, 3997.370800627546],
        [1490.4102994496432, 4000.0],
        [1400.0, 4023.918712198653],
        [1300.0, 4048.558469218259],
        [1200.0, 4071.305295092509],
        [1100.0, 4092.1748990692186],
        [1060.0124392467872, 4100.0],
        [1000.0, 4111.1848482834],
        [900.0, 4128.348190721229],
        [800.0, 4143.676046135198],
        [700.0, 4157.173923620604],
        [600.0, 4168.855459354523],
        [500.0, 4178.7279024480495],
        [400.0, 4186.796819535569],
        [300.0, 4193.060249381895],
        [200.0, 4197.525520724352],
        [110.21120510982271, 4200.0],
        [100.0, 4200.189296956492],
        [11.409296118577458, 4201.032313198807],
        [11.26971309299113, 4200.0],
        [9.859345855831664, 4100.0],
        [9.207088443836112, 4000.0],
        [8.701469586322228, 3900.0],
        [8.271085442199967, 3800.0],
        [7.888502586309943, 3700.0],
        [7.539590097895451, 3600.0],
        [7.216017876632443, 3500.0],
        [6.912321158587411, 3400.0],
        [6.624724552852864, 3300.0],
        [6.350477146505692, 3200.0],
        [6.087548114163271, 3100.0],
        [5.834345433212333, 3000.0],
        [5.589546958407851, 2900.0],
        [5.352131461752008, 2800.0],
        [5.121242119180592, 2700.0],
        [4.896265293719253, 2600.0],
        [4.676478944932127, 2500.0],
        [4.4614068208286914, 2400.0],
        [4.250628570151294, 2300.0],
        [4.043732413888003, 2200.0],
        [3.840343010283609, 2100.0],
        [3.640252131290516, 2000.0],
        [3.443099142331329, 1900.0],
        [3.2486450191940843, 1800.0],
        [3.0566671498447104, 1700.0],
        [2.8669554821101078, 1600.0],
        [2.6793182091171377, 1500.0],
        [2.4935476601936504, 1400.0],
        [2.309445776938012, 1300.0],
        [2.1268225766767004, 1200.0],
        [1.94549506194922, 1100.0],
        [1.7652930253117367, 1000.0],
        [1.586065699183146, 900.0],
        [1.4076821682836234, 800.0],
        [1.2300220890107458, 700.0],
        [1.052987058866334, 600.0],
        [0.8764933592021794, 500.0],
        [0.7004700317462764, 400.0],
        [0.5248567780708592, 300.0],
        [0.3496022369010007, 200.0],
        [0.1746625696061707, 100.0],
    ]
    hull = ConvexHull(boundaries_list)
    initial_df=pd.read_csv("sampled_points.csv", encoding="utf-8", sep=";")

    new_learner = LearnerND(some_f, hull)
    new_learner.load_dataframe(
        initial_df, with_default_function_args=False, point_names=("x", "y")
    )

    BlockingRunner(new_learner, npoints_goal=2110)

sampled_points.csv