0
What? It's problem isn't that the Spanish sample is too distant, because of "too Taforalt", it's because it's too close to use, it needs more distant samples like Belgian
Interesting, so you didn't see the Swiss French in the middle with 1.6 of distance?
Let me show that again to you:
Distance to: French_Provence-Alpes-Cote_d'Azur
0.01354540 Spanish_Mallorca
0.01408360 Italian_Aosta_Valley
0.01442538 Spanish_Penedes
0.01605360 Swiss_French
0.01610539 Spanish_Girona
Distance to: French_Marseille
0.01117725 French_Provence-Alpes-Cote_d'Azur
But yes, I agree that the French_South sample is a bad average, that's why I didn't use it.
Bookmarks