Predicting Weather Temperatures Using Sinusoidal Least-Squares Fit
4 次查看(过去 30 天)
显示 更早的评论
Hey Guys,
I'm in over my head in a coding class and I really need some direction with this problem. Basically I have weather temperature data from now, all the way back to 2013, taken everyday so each year is a [365x1] column vector (except for 2018 which is an [87x1]). I need to use a sinusoidal least-squares fit approach to predict weather 1, 3, and 5 years in the future. From my understanding of sinusoidal least-squares fit and Fourier analysis, I need to overlay a function over the real world data to create a model to predict weather. Again, I have a very limited understanding of this so please correct me if my thinking is incorrect. The function given in the problem that best represents the data is the picture I have included below. I have never inserted an equation this complicated into matlab before. So I need to run the temperature data through this function and use that to predict future weather temperatures, correct? If someone could help me out with this or provide a link to a generic code that would help me that would be great because I haven't made any progress at this point. Thanks in advance.
0 个评论
回答(1 个)
Chad Greene
2018-7-21
I wrote some functions including one called sinefit to do this for climate data. The three-parameter fit is as easy as
ft = sinefit(t,y)
where t is time, y is your temperature data, and ft contains the amplitude, day of maximum temperature, and mean value.
0 个评论
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Weather and Atmospheric Science 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!