There is also statistical function from scipy called stats.pearsonr() which basicly return two parameters correlation_coefficient, p_value = stats.pearsonr(x, y). correlation_coefficient is the same thing we see in df.corr() and a p-value is for statical test. A low p-value indicates that the correlation observed in the sample is unlikely to have occurred by random chance alone. for example with p-value = 0.05 we are 95% sure that there is a correlation between the variables.
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
What you also can do is a Lasso Regression and increase the alpha(lambda) from zero to x and then you will see which feature will fall to zero.
There is also statistical function from scipy called stats.pearsonr() which basicly return two parameters
correlation_coefficient, p_value = stats.pearsonr(x, y).
correlation_coefficient is the same thing we see in df.corr() and a p-value is for statical test.
A low p-value indicates that the correlation observed in the sample is unlikely to have occurred by random chance alone.
for example with p-value = 0.05 we are 95% sure that there is a correlation between the variables.
but the pearson correlation only takes correlation with two variables i think
The problem with corrleation is that it only will detect collinearity but wont detect multi-collinearity...