Shade standard deviation for time series

Question

JAIME DIEGO RICO 2020-7-10

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/562826-shade-standard-deviation-for-time-series

评论： Star Strider 2020-7-18

采纳的回答： Star Strider

在 MATLAB Online 中打开

Hi!

I'm trying to shade the std (daily_std) for mean sst (daily_average).

I've seen similar questions, but those codes don't work for me.

My x-axis is one year of date times. My y axis is the mean of sst.

I've done the following code:

figure
curve1=daily_average(:,1)+daily_std(:,1); %only column 1 because every column is one different location
curve2=daily_average(:,1)-daily_std(:,1); 
plot(date,curve1,'r')
hold on
plot(date,curve2,'r')
hold on
date2 = [date, fliplr(date)];
inBetween = [curve1, fliplr(curve2)];
fill(date2, inBetween, 'g');
hold on;
plot(date, daily_average(:,1));

However, I obtain the following result:

Any idea?

Thank you very much

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

Star Strider 2020-7-10

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/562826-shade-standard-deviation-for-time-series#answer_464036

在 MATLAB Online 中打开

I cannot run your code, so I cannot test this.

However, fill may not be appropriate here. See if using patch instead does shat you want:

patch(date2, inBetween, 'g');

Your curtrent syntax will work for patch, so no other changes are necessary.

5 个评论
显示 3更早的评论隐藏 3更早的评论

Star Strider 2020-7-10

编辑：Star Strider 2020-7-10

在 MATLAB Online 中打开

The only option appears to be to convert the datetime array to date numbers for the plot, then use patch.

Example —

dt = datetime('now')+calmonths(1:10);
dn = datenum(dt);
upper = rand(1,10)+2;
lower = rand(1,10);
figure
patch([dn fliplr(dn)], [upper fliplr(lower)], 'g')
datetick('x', 'mmm', 'keepticks','keeplimits')

EDIT — (10 Jul 2020 at 17:39)

This works, and produces the correct result with datetime vectors with my code:

figure
fill([dt, fliplr(dt)], [upper, fliplr(lower)].', 'g')
grid

so with respect to your code, try this:

fill(date2, inBetween.', 'g');

I rarely use fill, preferring patch, however it seems that leaving the first argument (independent variable vector) as a row vector and transposing the second argument (dependent variable vector) to a column vector will do what you want with fill. This is not in the documentation (at least I cannot find any reference to it), and simply required a bit of experimentation for me to discover it.

It produces the same result as the patch call with a datenum vector, without the inconvenience of having to convert the datetime vector to datenum.

.

JAIME DIEGO RICO 2020-7-18