Setting up Quality Control on dataset

Question

Sara 2011-2-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1313-setting-up-quality-control-on-dataset

Hi,

I am trying to write a quality control program for a large dataset. I need to check for three conditions: 1) If values are outside a certain limit (eg 0-1000), 2) If there is too large a jump between consecutive values (eg >250) 3) If there is no change of values within a certain time frame (eg 4 hours (each row will be timestamped))

I have absolutely no idea how to write this, except that I have a vague idea of writing a loop and a list of nested if statements to create a second array of 0 (no errors) 1, 2 or 3. But I don't know how to compare subsequent values or check blocks of data for any change.

I'm not asking for someone to write my code for me, but does anybody have any ideas?

Thanks,

S.

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

Walter Roberson 2011-2-14

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1313-setting-up-quality-control-on-dataset#answer_1901

在 MATLAB Online 中打开

out_of_range = V < 0 | D > 1000
if any(out_of_range)
   ....
end
diffs = [diff(V) 0];
bigjumps = abs(diffs) > 250
if any(bigjumps)
   ...
end
ldiff = [true ~logical(diffs)];
runstarts = strfind(ldiff, [1 0]);
runends = strfind(ldiff, [0 1]);
runlengths = runends - runstarts;
if any(runlengths > ...)

note that the code for detecting the runs of identical values would have to be adjusted if the sampling is not at even intervals, as then number of samples in the run would not translate directly to time of the run.

You also need to take in to account the possibility that there is noise in the sampling system, and thus that you might want to count a difference of less than some tolerance as being "the same value".