Renewable energy sources have increasingly been integrated into power systems because of their superior environmental, technical, and economic advantages. However, the uncertainty of these sources besides load fluctuations and equipment outages brings new challenges to power system operating and planning studies. Also, the correlation between uncertain variables makes operational decisions more and more complicated. Probabilistic assessment of power systems is essential to evaluate uncertainties' effects and make reasonable decisions. Methods with high accuracy and low computational burden are efficient for online and fast requirements such as optimal power flow (OPF) problems. A clustering method based on the K-medoids technique is used for the probabilistic assessment of the power system in the OPF problem. Unlike other clustering techniques the K-medoids method can manage discrete variables such as equipment outages besides other continuous variables such as load and wind generation. This paper presents an OPF problem considering the uncertainty of renewable energy sources, load fluctuations, and transmission lines outage as well as the correlation among them. The probabilistic assessment is conducted by the K-medoids method and the optimization problem is solved by the cooperation search algorithm (CSA) as an efficient evolutionary algorithm. Numerical results using the IEEE standard 30 and 118 bus test systems show the efficiency of the proposed method in terms of accuracy and computational burden. This paper solves the probabilistic optimal power flow considering the uncertainty of renewable energies, loads, and component outages as well as the correlation among uncertain variables. The k_medoids data clustering technique has been used for the simultaneous clustering of discrete and continuous variables in uncertainty management. Also,the cooperation search algorithm (CSA) is used for solving the optimal power flow problem.image