Update README.md

Treaps/README.md

@@ -4,3 +4,15 @@ Each node in this tree has a pair __(X,Y)__, Binary search tree property gets sa
 and right subtree elelments > X.
 Also for Y all nodes in left and right subtree are less than Y.Y values are called **priority**.
 This data structure is often useful in balancing BST. If priorities are chosen randomly height of this binary search tree is **O(log N)** 
+
+As per Insert procedure , key point is anything you insert left subtree contain node < F and right subtree > F
+- P(F) < P(J) we cannot insert here, since K(F) < K(J), goto left subtree.
+- P(F) > P(C) , so we know this is the right place of F , call split ( C, F, F->L, F->R).
+- K(F) > K(C) so we know that **F->L = C** but we still have to find F->R , which would be __first maximumum__ found in **right subtree**.
+  split(C->R, F, C->R, R)
+- K(G) > K(F), **first maximum** found , this right subtree of **F->R = G**, now **C->R which was earlier G**, is changed, so find **first minimum** in **left subtree** i.e. G->L split(G->L, F, L, G->L)
+- K(D)< K(F), we found first minimum and this will be added to **C->R = D** , once again **G->L** changed , so find first maximum in right subtree.
+
+
+So recurively this process repeats until we reach empty nodes.
+