Hi, everyone!
I am not expert in wine. Well, I am afraid that becoming expert in wine has risks of alcohol addiction. Moreover, I am a poor PhD student who has no money. Nevertheless, I enjoy drinking wine with my girlfriend and watching films.
Very skewed distribution for the price. Well, it is usually what happens with price - there are few elite wines for and people who have enough money to pay for that.
Let’s have a look on logarithmized price distribution from several countries:
Distribution of points for wines is much more symmetrical and not far from normal:
The most obvious question is whether wine price and its quality (as measured by some nerd experts) are correlated. Well, of course their reviews are not blind to the price and, well, more expensive wine is, in general, better.
Non-parametric correlation:
##
## Spearman's rank correlation rho
##
## data: points and price
## S = 4.7357e+13, p-value < 2.2e-16
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
## rho
## 0.5986811
…and Pearson correlation of points with logarithmized price:
##
## Pearson's product-moment correlation
##
## data: points and log(price)
## t = 230.36, df = 89126, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
## 0.6067641 0.6149944
## sample estimates:
## cor
## 0.6108958
Correlation is rather strong, and it is not a big deal. Let’s plot it.
It is not reasonable to plot every country - there are so many countries with only few wines:
I live in Russia (fortunately, no Russian wines in the list) and, maybe, wines from neiborgh countries would be easier to get. I can do scatterplot for them:How to get the best price for a few euros from my pocket. Remember, I am a human being (and student!), wine is not the only thing that I consume. I calculate quality per logarithmized ratio and created a joy plot for that: