I'm regressing "diamond price" which is a dependent variable on 10 independent variables. The regression shows a significant p-value for ANOVA, R-squared and Adjected R-square are above 0.67. P-values for all the independent variables are significant. But correlation between the dependent variable and 8 of my independent variables are close to zero, also scatterplots shows nonlinear relationship, all the scatterplots show similar pattern. I'll upload one as an example. Besides, the distribution of my dependent variable is highly skewed to the right. Can I transform both the dependent variable and the 8 of my independent variables at the same time? How do I do that?

Table percentage of dimaond vs total_sales price