related analysis
Two variables, X, Y
| Variable X | Variable Y | Correlation |
|---|---|---|
| Increases | Increases | Positive correlation |
| Increases | Reduced | Negative correlation |
| Increases | Ignore | irrelevant |
Seeking association
协方差
方差Is used to measure the extent of a single random discrete variable
协方差is generally used to characterize the degree of similarity of two random variables
| Covariance> 0 | X, Y positive correlation |
|---|---|
| Covariance <0 | X, Y negative correlation |
| Covariance = 0 | X, Y independent and unrelated |
pearson相关系数
r in the range [-1, 1]
| r value | relativity |
|---|---|
| 0.8-1.0 | Extremely relevant |
| 0.6-0.8 | Strong correlation |
| 0.4-0.6 | Moderate related |
| 0.2-0.4 | Weak correlation |
| 0-0.2 | No |
Correlation analysis function
| method | Explanation |
|---|---|
| .These () | Covariance matrix |
| .corr() | Calculating a correlation coefficient matrix, Pearson, Spearman, Kendall et coefficient |
