[ΑΝΝ] WIND FORECASTING WITH A LOT OF MISSING DATA (MATLAB OR PYTHON)

Dear Sir/ Madam, At first I hope you are all fine. I have some data from some aeolic parks (wind parks) and I need to get good results for a forecast. I have done most of the preproccessing. I will send you in chat the full explanation in a doc file, the code I have done so far and the data so you can check. I have 2 problems, a small one and a bigger one, I will need help with: 1. I missed some of the matlab code. At the code I developed, I have Identified the missing data, replaced it with possible linear relation, verified to the maximum extent with wind speed, recognized the trend and correlated with farm data. Also I have done a simple algorithm to add the correct date and time stamps to my excel file because this was a problem with the data. 2. The bigger problem: Data after linear fitting were not sitting on the normal curve. Fitted data is out of trend. So Wind modelling was useless in this case. As in many cases two columns (nacelle position and temperature) were also missing. If data sits on the trend line, then neural network can easily predict the future trend. Missing data in Matlab can easily be filled using using command "ismissing" and then filling the missing data using the "fillmissing" command of matlab. Most convenient is linear or spline. Because entries are in range of 21000, so searching and filling alone killed the freelancers RAM and crashed. I have 16gb RAM so you can use it if you think it is enough . In any case I will be happy to answer your questions. I can send you (some of the) the preprocessing code, the complete requirement, and the data in chat so you can check them thoroughly. Your mission is to fix the data, make the results realistic and fullfill the requirements. The budget is $150 and we have 4 days. Thank you a lot