How to convert decimal hour and calculate the average every 5 min?

Hello all! I'm a beginner at Matlab and I need to solve a problem. First, I need to convert the UT columm from decimal hour to hour:minute:seconds. Then, I need to calculate every 5 minutes on average the ROT collumn and show the reply in a new matrix (hour:min:sec,rot mean).

Example:

e.g. UT (5.404)=0.404*60=24.252; 0.252*60=15.12 ,then UT(5.404)=5:24:15 hours:min:sec

Thanks in advance

Marcelo

[Doug Hull]

Why are you dropping the 0.004 in the first step?

>> [Y, M, D, H, MN, S] = datevec(5.404/24)

Y =

       0

M =

       0

D =

       0

H =

       5

MN =

      24

S =

     14.4000

[Marcelo Duarte]

Thank you very much Doug! Now, how can I show this result at 5:24 format at UT column together with the ROT column?

Thanks!

Answer 1

[Walter Roberson]

sprintf('%02d:%02d', H, MN)

In order to show this "together" with the ROT, you will need to indicate the mechanism you wish to use to "show" it. Are you referring to a uitable() for example?

Ordinary numeric matrices cannot hold a combination of strings and numbers.

[Marcelo Duarte]

Thanks Walter Roberson, I used textscan command to read dat file and stored as ut=data{1} and rot=data {2}.I need to calculate every 5 minutes about this equation:

ROTI=sqrt(<rot^2>-<rot>^2)

then I did

1. rot_quad=rot.^2;
2. rot_quad_mean=mean(rot_quad);
3. rot_mean=mean(rot)
4. rot_mean_quad=rot_mean^2
5. ROTI=sqrt((rot_quad_mean-rot_mean_quad)). %e.g: 5:24 to 5:29= step 1 to 4

and finally ROTI. Thanks in advance!

[Walter Roberson]

bin_num = 1 + floor(ut(:) * 60 / 5) - floor(ut(1) * 60 / 5);
max_bin_num = max(bin_num);

rot_quad = rot.^2;
rot_quad_mean5 = accumarray(bin_num, rot_quad(:), [], @mean);
rot_mean5 = accumarray(bin_num, rot(:), [], @mean);
rot_mean_quad5 = rot_mean5 .^ 2;
ROTI_5 = sqrt(rot_quad_mean5 - rot_mean_quad5);

first_date_vec = datevec(ut(1) * 24);
first_date_hour = first_date_vec(4);
first_date_min5 = floor(first_date_vec(5) ./ 5) .* 5;
first_date_ser = datenum([0 0 0 first_date_hour first_date_min5 0]);
bin_date_vecs = datevec(addtodate(first_date_ser, 5 * (0:max_bin_num-1), 'minute'));
ROTI_matrix = [bin_date_vecs(:,4), bin_date_vecs(:,5), ROTI_5];

Now, ROTI_matrix will be an array, first column the hour of each 5 minute interval, second column the minute of each 5 minute interval, third column the corresponding ROTI for the 5 minute interval.

[Marcelo Duarte]

Hi Walter! When I try to run the code, appears an error at addtodate (D,Q,F) command: addtodate (line 49) Quantity must be a numeric scalar. Then, I added the "norm" command in the Q field for convert from vector into scalar number:

 bin_date_vecs = datevec(addtodate(first_date_ser, norm( 5 * (0:max_bin_num-1)), 'minute'));

Is it correct do this change?

After the change, I tried to run the code again but, appeared other error: Error using horzcat Dimensions of matrices being concatenated are not consistent.

Error in roti2 (line 40) ROTI_matrix= [bin_date_vecs(:,4), bin_date_vecs(:,5), ROTI_5];

How can I resolve the error horzcat?

Thanks in advance!

Answer 2

[Andrei Bobrov]

k = rem(UT,1)*60;
tms = fix([UT,k]);
m = ceil(k/5)*5;
[~,b,ii] = unique(m);
out = [tms(b,:), accumarray(ii,ROT,[],@mean)];

[Marcelo Duarte]

Thanks Andrei! The program generated a 'b' array 13x1 and another 'out' array 13x3. I used a file with 5 hours of data, so 'b' and 'out' arrays needed to have 61 average values. Can you help me solve it? Thanks in advance.