Averaging data every three hours

Hi everyone!

I have been looking for one solution for this issue but with no results.

From the next data set:

YEAR  MONTH  DAY  HOUR  MINUTE  SECOND  WIND DIRECTION (DEGREES)
2009  1  1  0  0  0  20
2009  1  1  0  10  0  21
2009  1  1  0  20  0  24
2009  1  1  0  30  0  29
2009  1  1  0  40  0  30
2009  1  1  0  50  0  28
2009  1  1  1  0  0  24
2009  1  1  1  10  0  31
2009  1  1  1  20  0  27
2009  1  1  1  30  0  25
2009  1  1  1  40  0  29
2009  1  1  1  50  0  31
2009  1  1  2  0  0  30
2009  1  1  2  10  0  30
2009  1  1  2  20  0  25
2009  1  1  2  30  0  27
2009  1  1  2  40  0  18
2009  1  1  2  50  0  27
2009  1  1  3  0  0  27
2009  1  1  3  10  0  29
2009  1  1  3  20  0  29
2009  1  1  3  30  0  30
2009  1  1  3  40  0  30
2009  1  1  3  50  0  30
2009  1  1  4  0  0  34
2009  1  1  4  10  0  38
2009  1  1  4  20  0  43
2009  1  1  4  30  0  57
2009  1  1  4  40  0  79
2009  1  1  4  50  0  72
2009  1  1  5  0  0  65
2009  1  1  5  10  0  64
2009  1  1  5  20  0  59
2009  1  1  5  30  0  56
2009  1  1  5  40  0  51
2009  1  1  5  50  0  51
2009  1  1  6  0  0  67
2009  1  1  6  10  0  60
2009  1  1  6  20  0  61
2009  1  1  6  30  0  78
2009  1  1  6  40  0  82
2009  1  1  6  50  0  88
2009  1  1  7  0  0  87
2009  1  1  7  10  0  85
2009  1  1  7  20  0  96
2009  1  1  7  30  0  92
2009  1  1  7  40  0  99
2009  1  1  7  50  0  108
2009  1  1  8  0  0  123
2009  1  1  8  10  0  118
2009  1  1  8  20  0  101
2009  1  1  8  30  0  99
2009  1  1  8  40  0  108
2009  1  1  8  50  0  105
2009  1  1  9  0  0  104
2009  1  1  9  10  0  105
2009  1  1  9  20  0  101
2009  1  1  9  30  0  95
2009  1  1  9  40  0  96
2009  1  1  9  50  0  101
2009  1  1  10  0  0  96
2009  1  1  10  10  0  113
2009  1  1  10  20  0  126
2009  1  1  10  30  0  107
2009  1  1  10  40  0  105
2009  1  1  10  50  0  88
2009  1  1  11  0  0  86
2009  1  1  11  10  0  72
2009  1  1  11  20  0  72
2009  1  1  11  30  0  70
2009  1  1  11  40  0  79
2009  1  1  11  50  0  81
2009  1  1  12  0  0  90
2009  1  1  12  10  0  101
2009  1  1  12  20  0  106
2009  1  1  12  30  0  100
2009  1  1  12  40  0  102
2009  1  1  12  50  0  93
2009  1  1  13  0  0  96
2009  1  1  13  10  0  91
2009  1  1  13  20  0  97
2009  1  1  13  30  0  100
2009  1  1  13  40  0  94
2009  1  1  13  50  0  101
2009  1  1  14  0  0  86
2009  1  1  14  10  0  86
2009  1  1  14  20  0  88
2009  1  1  14  30  0  87
2009  1  1  14  40  0  93
2009  1  1  14  50  0  98
2009  1  1  15  0  0  92
2009  1  1  15  10  0  92
2009  1  1  15  20  0  79
2009  1  1  15  30  0  79
2009  1  1  15  40  0  85
2009  1  1  15  50  0  89
2009  1  1  16  0  0  95
2009  1  1  16  10  0  84
2009  1  1  16  20  0  79
2009  1  1  16  30  0  75
2009  1  1  16  40  0  86
2009  1  1  16  50  0  79
2009  1  1  17  0  0  97
2009  1  1  17  10  0  84
2009  1  1  17  20  0  67
2009  1  1  17  30  0  72
2009  1  1  17  40  0  80
2009  1  1  17  50  0  89
2009  1  1  18  0  0  106
2009  1  1  18  10  0  21
2009  1  1  18  20  0  29
2009  1  1  18  30  0  320
2009  1  1  18  40  0  315
2009  1  1  18  50  0  319
2009  1  1  19  0  0  325
2009  1  1  19  10  0  323
2009  1  1  19  20  0  330
2009  1  1  19  30  0  332
2009  1  1  19  40  0  325
2009  1  1  19  50  0  328
2009  1  1  20  0  0  330
2009  1  1  20  10  0  328
2009  1  1  20  20  0  333
2009  1  1  20  30  0  331
2009  1  1  20  40  0  341
2009  1  1  20  50  0  338
2009  1  1  21  0  0  340
2009  1  1  21  10  0  341
2009  1  1  21  20  0  350
2009  1  1  21  30  0  348
2009  1  1  21  40  0  354
2009  1  1  21  50  0  350
2009  1  1  22  0  0  358
2009  1  1  22  10  0  353
2009  1  1  22  20  0  359
2009  1  1  22  30  0  354
2009  1  1  22  40  0  2
2009  1  1  22  50  0  8
2009  1  1  23  0  0  3
2009  1  1  23  10  0  14
2009  1  1  23  20  0  13
2009  1  1  23  30  0  15
2009  1  1  23  40  0  20
2009  1  1  23  50  0  17
2009  1  2  0  0  0  25

It is necessary an average of data (wind speed) exactly every three hours

The results desired after the application of the filter are the next:

YEAR  MONTH  DAY  HOUR  MINUTE  SECOND  Wind speed (average three hours)
2009  1  1  3  0  0  26.47
2009  1  1  6  0  0  49.11
2009  1  1  9  0  0  94.11
2009  1  1  12  0  0  93.50
2009  1  1  15  0  0  95.06
2009  1  1  18  0  0  84.28
2009  1  1  21  0  0  294.89
2009  1  2  0  0  0  182.44

Best regards

Answer 1

[Jan]

Hour3 = ceil(Hour / 3) + 1;
Mean3 = splitapply(@mean, WindSpeed, Hour3);

I assume it is a typo, that you want to calculate the wind speed based on the "WIND DIRECTION (DEGREES)". It is not clear how your data are represented, because the shown table might be a file, list of vectors or a table object. I dared to write only "Hour" and "WindSpeed" and leave it up to you, how these variables are created.

The shown data look like the minutes are not relevant. If this is not true, consider than by adding them to the Hour value before building the groups.

Answer 2

[dpb]

Use Matlab internal storage order to do this over regularly-fixed data...

>> wdir=importdata('luis.txt');   % the data array
>> mean(reshape(wdir(1:end-1,end),18,[])).'
ans =
 26.4444
 46.8889
 92.0556
 94.2778
 94.9444
 83.5000
281.8889
199.9444
>>

This doesn't quite match your numeric value(s); I can get the first value above if I include the 0-hour of the next day in the first average but can't reproduce the rest; not sure what you did...

>> mean(wdir(1:19,end))  % this matches but last point is 0 hr of day 3, _not_ day 2
ans =
 26.4737
>> mean(wdir(18+(1:19),end))  % did average from 19th point???  Apparently not...
ans =
 47.9474
>> mean(wdir(18:36,end))  % well, 19 points from 18th??? nope, not that, either
ans =
 45.8421
>>

So, don't know where the others came from after the first, but I think the correct answer should be that given above.

NB: The column header is "Wind Direction" not "Wind Speed" and the numbers pretty-much indicate it has to be that...might want to fix up your output title to match.