Building A Pricing Model For Diamonds

Building a pricing model for diamonds.

In order to build a model first we want to see a relationship among our variables. First, I made a matrix of plots.

When I have two variables that are normally distributed I try to see their relationship. When I run the plot in R, I can see that we get a linear regression. The correlation price on carat is very high almost 0, 97.

Then I go to see if color has any relationship in our model. Therefore, I ask R to color all the dots depending on their color. In addition, I see a clear distribution of different diamonds and color. High quality color always is more expensive that low quality color of the same size stone.

Towards the end distribution of color, J is wider. From this, I conclude that people prefer size to color. If they buy a large stone, it is more important for them that it would be big rather than of good color. However, when people buy small stone, they want to compensate the size with other factor that increase the value of the stone. If a person buys very small diamonds, he often choose the best color.