from sklearn.manifold import MDS
from sklearn.decomposition import PCA
from sklearn.manifold import TSNE
import urllib.request
from gensim.models import Word2Vec

sonad=urllib.request.urlopen("https://minitorn.tlu.ee/~jaagup/oma/too/23/05/word2vec/sonaindeksid.txt").read().decode("utf-8").split("\n")
sonad=[rida.split()[0] for rida in sonad if rida]
print("sinad")
mudel=Word2Vec.load("https://minitorn.tlu.ee/~jaagup/oma/too/23/05/word2vec/lemmad1.model")
print("loetud")
kohad2=PCA(n_components=5).fit_transform(mudel.wv.get_normed_vectors())
print(kohad2)
#kohad3=MDS().fit_transform(kohad2)
kohad3=TSNE().fit_transform(kohad2)
print(kohad3)
vastus=[rida[0]+","+str(rida[1][0])+","+str(rida[1][1]) for rida in zip(sonad, kohad3)]
f2=open("paigutus1.txt", "w", encoding="utf-8")
print("\n".join(vastus), file=f2)
f2.close()