Z-scores of time series data (temperatures)

Question

Tyler Smith 2017-8-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/353109-z-scores-of-time-series-data-temperatures

评论： Chad Greene 2017-8-17

I need to compute the z-scores of daily temperature data. My data runs from 1975 - 2010, November - March. Each November - March is relates to each day of the season's average temperature (1 - 151 days of season, so 151 avg temperatures). For example, a z-score for day 1 of the season (November 1st) would begin with an average for all 35 Nov. 1st days (1975 - 2010). I computed this in Excel by fitting a polynomial regression to each day of the season's average (151 averages), then divided the output of the polynomial regression equation for each day by the standard deviation of that particular day, but this is quite taxing to compute over multiple large data sets. Any ideas on how to approach this would be great!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Chad Greene 2017-8-17

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/353109-z-scores-of-time-series-data-temperatures#answer_278302

Does the zscore function do what you want?

2 个评论
显示无隐藏无

Tyler Smith 2017-8-17

I don't believe so. If I'm not mistaken, the zscore function computes the zscore of all temperatures without respect to the day of the season. The day of the season is a necessary component because climatologically each day of the season has a different avg, thus a different stdev.

Chad Greene 2017-8-17

It sounds like you could loop through 1 to 151 and compute the z scores for each day? Alternatively, you could deseason the signal and then calculate the z scores for the whole dataset.

请先登录，再进行评论。