A single data point like that should never be considered anything close to "absolute proof" of anything - because you have absolutely no way of knowing that, fo

A single data point like that should never be considered anything close to "absolute proof" of anything - because you have absolutely no way of knowing that, for whatever reason (random chance, or the food you were eating at the time, or a compliment somebody paid you on the day you started taking them, or....), you might have felt better on those first days even if you hadn't started taking antidepressants at all. Correlation is not causation, as they say. (Hope your depression is gone or at least not too bad now days, regardless of what drugs or placebos may have played a part!)