how to delete the first n data entries from each column of a vector?

I have a large vector of data (1600 x 240), and I was wondering if anyone knows how to delete the first n number of entries from each column of this vector.

As the data entries do not start at the same date for each column I cannot simply delete the first n rows, as for some columns the data has not started by then.

thanks in advance

Answer 1

[Star Strider]

I’m not certain exactly what you want.

See if this works:

Matrix = rand(5);                           % Create Data
n = 2;                                      % Assign ‘n’
Matrix(1:n,:) = [];                         % Result (Delete First ‘n’ Rows From All Columns)

[Daniel Hooker]

Hi thanks for your help.

The problem is that what you suggests deletes the first n rows from the whole dataset, however what I am trying to achieve is to delete the first 36 data entries from each column. The tricky part is that the data entries do not start in the same row for each column (i have attached a screenshot of the excel document I am importing if that makes it any clearer)

thanks again

[Star Strider]

When xlsread imports the file, the empty elements will be read as NaN. The constraint here is that the size of the matrix must not change, so the only way to do what you want is to set the rows in each column that you want to remove to NaN as well. You can then deal with the NaN values (that you will have to deal with regardless) any way you want.

I can’t find a way to do this without a loop. This is the most efficient way I can come up with to do what you want.

The Code —

A = [NaN  NaN  NaN  1  NaN 2  3;  NaN  NaN  NaN  1  NaN 2  3;  NaN  4  NaN  1  NaN  2  3;  NaN  20  NaN  1  17 2  3; 6 7 8 9 10 11 12;  randi(50, 4, 7)]; 
Anan = ~isnan(A);                                                           % Logical Vector
[Ananr,Ananc] = find(Anan);                                                 % Indices Of ‘~isnan’
StartRow = accumarray(Ananc, Ananr, [], @min);                              % First ‘non-NaN’ Row
RemoveRows = 2;                                                             % Rows To Remove
Anew = A;                                                                   % Retain Reference
for k1 = 1:length(StartRow)
    IdxRng = StartRow(k1):StartRow(k1)+RemoveRows-1;
    Anew(IdxRng,k1) = NaN;                                                  % Set Removed Rows To ‘NaN’
end

produces:

A =
   NaN   NaN   NaN     1   NaN     2     3
   NaN   NaN   NaN     1   NaN     2     3
   NaN     4   NaN     1   NaN     2     3
   NaN    20   NaN     1    17     2     3
     6     7     8     9    10    11    12
    21    22     6    17    37    48    36
    33     1    19    48    14    21    34
    32    50    10    47    22    50    27
    15     9    25     3    28    16    35

Anew =
   NaN   NaN   NaN   NaN   NaN   NaN   NaN
   NaN   NaN   NaN   NaN   NaN   NaN   NaN
   NaN   NaN   NaN     1   NaN     2     3
   NaN   NaN   NaN     1   NaN     2     3
   NaN     7   NaN     9   NaN    11    12
   NaN    22   NaN    17    37    48    36
    33     1    19    48    14    21    34
    32    50    10    47    22    50    27
    15     9    25     3    28    16    35

The code first produces a separate logical index array (not necessary, it can be combined in the find call, I did it for diagnostic purposes), then find the minimum start indices of the rows for each column with the accumarray call. It then loops through those and replaces the ‘RemmoveRows’ number of rows in each row with NaN values, effectively eliminating them.

Answer 2

[Image Analyst]

Try this:

Ri_Rf = Ri_Rf(n+1:end, :); % Extract from rows (n+1) until the bottom.

[Daniel Hooker]

Hi thanks for your help.

The problem is that what you suggests deletes the first n rows from the whole dataset, however what I am trying to achieve is to delete the first 36 data entries from each column. The tricky part is that the data entries do not start in the same row for each column (i have attached a screenshot of the excel document I am importing if that makes it any clearer)

thanks again