How to config parameters to match sklearn.cluster.MeanShift

[JohnPekl]

Thanks for your great work.

Could you help me configure parameters in your code to match sklearn.cluster.MeanShift scikit-learn 0.21.1 output?

from sklearn.cluster import MeanShift

data = np.loadtxt("test.csv", delimiter=",")
cluster = MeanShift(bandwidth=1).fit(data)

This code outputs 30 clusters. However, your C++ outputs 13 clusters (with the same float bandwidth = 1.0;). If I set bandwidth=0.45 in your code, it outputs 30 clusters but the cluster center of sklean and your code are not identical.

I think, it is better to set clusterEps scalable to bandwidth rather than fix it in this line.

Number of points: 85
Number of dimensions: 2
Number of clusters: 13
Elapsed time: 0.0038222 s

Data is given below.

3.194799644708685271e+00,1.399483126857062487e+01
3.658620824564967400e+00,7.998179031752796320e+00
2.406295091169551537e+00,1.371293923319686137e+01
6.455143433800709651e+00,1.332078919555413599e+01
6.639544003222972002e+00,-9.784779332706614596e-01
4.683512497858520085e+00,1.533389839307644564e+01
3.030875878761376985e+00,7.453306804175237943e+00
5.434090247273947405e+00,1.414049843741694978e+01
3.679344315396408938e+00,1.467462112161013366e+01
7.768518143283098532e+00,1.332199924661608215e+01
-2.319508287202100227e+00,2.477186757495290603e+00
1.074944778405868995e+00,9.340254026825610012e+00
6.894037470483292296e+00,7.581186678942872237e-02
4.844635093320603936e+00,1.203458719558798506e+01
-2.795300513931746611e+00,2.471883609154136963e+01
7.915733620177003438e-01,1.721061193132751299e+01
7.132313391239341271e+00,1.550459686982297214e+01
8.515816767482528249e+00,1.065819225023263783e+01
8.978747615487289169e-01,1.043591726019572619e+01
5.546071571953522295e+00,1.562121487999592873e+01
4.065270136084711439e+00,1.583814811578007919e+01
3.184036959812714535e-02,1.459321007541508486e+01
-1.244543671389629180e+00,1.743108984169777287e+01
6.390077321532942278e+00,6.840997324649358280e-01
8.147730609299953741e+00,1.365558937915167270e+01
6.047264590048973432e+00,1.368315895680113492e+01
-2.237954885311399078e+00,1.097428962799293295e+01
4.112337877847533463e+00,1.498296194188718644e+01
6.093217411552747542e+00,1.521741446097545669e+01
3.519516426117108576e+00,1.717652937955963210e+01
7.717618815028292723e+00,1.418862816020948969e+01
7.072528949742961091e-01,1.730178780786865644e+01
6.956095223985222731e+00,1.571501162530448958e+01
2.886951082285038162e+00,1.447572078353684866e+01
2.435018839978735805e+00,1.643058716302973465e+01
5.557089553165287299e+00,1.500259061064638999e+01
1.966322507606136361e+00,1.471772285493714350e+01
3.591336047378127727e+00,1.466238736293894718e+01
8.106564362560180204e-01,1.022956186631088826e+01
2.730131991988756557e+00,8.081727808511182332e+00
6.753675108425203355e+00,1.714773673301615631e+01
4.860962701368696237e+00,1.501714466415207205e+01
3.461255980400441246e+00,8.426995794166828091e+00
5.089684241933406739e+00,4.213103664488715872e+00
6.459589772973727939e+00,1.351150064898626724e+01
3.682710121593864905e+00,7.784238822473763975e+00
6.519441637177784798e+00,1.751059074617726363e+00
8.519958784130727381e+00,1.063034061040542433e+01
-2.271280185668279383e+00,3.879749441538207666e+00
-2.782982246156696871e+00,2.261527601462014925e+00
6.848669350998769723e+00,-3.300827752221471845e-01
-1.947588058896889684e+00,-3.393087653162231709e-01
-2.191119170120348958e+00,-2.343437401262641107e+00
7.712764907854805507e+00,1.319482344711862432e+01
5.797760047057320065e+00,1.509748037202986382e+01
-2.053737638193722237e+00,1.750538842583448940e+00
8.122384948222315160e+00,1.379717402367879231e+01
-2.545057085277153242e+00,-3.696450699828900399e-01
8.778814720324358278e-01,9.700049742821180487e+00
4.693773419951842385e+00,1.236804175668522809e+01
6.603024160919162000e+00,1.519063647432190933e+01
6.402927408401876797e+00,-6.310552586436059208e+00
5.448012673099911218e+00,1.449952116547625280e+01
-1.946289969373644624e+00,-4.386493878635892574e-01
-2.290784283403446420e+00,-1.068402841612288823e+00
-2.145998628910888328e+00,-2.420282125602403411e+00
-2.210651295857830156e+00,2.539023175221418516e+00
-2.368403256205440677e+00,-1.843572973551060423e+00
-2.313993176597985535e+00,1.113922159618876506e+01
5.304183496810864540e+00,1.597518979504664260e+00
7.358031045789820146e-01,9.957558811225338147e+00
2.320085637716873261e+00,7.920250745862526642e+00
1.921114181617376238e+00,1.656865295717014064e+01
8.308942694396936446e+00,7.175072308496275308e-01
8.290689221475529891e+00,3.117030130446184266e+00
5.969739229591654706e+00,1.490792651282334980e+01
4.392748810510084922e+00,2.552749483389695939e+01
4.424134938057694200e+00,1.361148649320027637e+01
-2.016840179546761647e+00,9.798201658621681354e-01
3.835888263461657122e+00,7.994134963339781130e+00
-6.620268053846491974e-01,1.697437277727217975e+01
2.967707110665201764e+00,7.958832455032286646e+00
-2.193787919297927047e+00,3.106307822680461239e+00
1.022181997007836962e+00,1.038669874254246928e+01
2.674588284866549426e+00,1.395835657759778314e+01