Check if Column has Correct Sequence of Numbers

I have a column of motion tracking data where there are supposed to only be 3 values (integers) and they should repeat in the same order every 3 rows.

For example, the order is 52630, 1, 2. It repeats ilike that for a few thousand lines.

Once or twice in the column, it will skip one of these values. I need to be able to detect when this happens, and if possible remove the rows from the table where it happens.

For example, if there is a 1 missing in the sequence, I need to be able to delete the rows that containt the 52630 and 2 that are missing the 1 in between them.

I've tried a few different things but none seemed to work as I expected them to.

Thanks!

Answer 1

[KALYAN ACHARJYA]

Steps:

1.Create the complete Table with original data (including missing data)

2. use ismissing function to detect in Table

3. Get the rows from the logical matrix of the result of step 2

[r,~]=find(result3==1);

4. Remove all those rows from the table

table_varible_name(r,:)=[]

[Spencer Ferris]

Ah I think I didn't explain exactly what I meant clearly. There are no cells in the table that are empty, see example table below. It just skips from the 52630 to the 2. In this situation I would want to remove rows 7 + 8.

[KALYAN ACHARJYA]

As per the given attached data here an example

data=repmat([52630 1 2],[1,5]);
data(8)=[]; % Creation of sample data
T=table(data')

T =

14×1 table
    Var1 
    _____
    52630
        1
        2
    52630
        1
        2
    52630
        2
    52630
        1
        2
    52630
        1
        2

First Option: (Hard Coding)

result=diff(T.Var1)

Here you have to make a comparison in the result data or use ~"find" function to those undesired results, as you have known patterns, you know the diff results of consecutive numbers.

result =
      -52629
           1
       52628
      -52629
           1
       52628
      -52628
       52628
      -52629
           1
       52628
      -52629
           1 

See index 7 number -52628 is not matched with pattern diff results.

Second Option

Read about Loren Blog Here (Pattern Matching in the array)

idx = findpattern(T.Var1,[52630 1 2])

idx =

1 4 9 12

idx_final=sort([idx,idx+1,idx+2]); % This you can generelize; 
% it gives the 1st index of the pattern matched, three number in the pattern 
% Hence I have added extra 2 to get the indices
% Total addition required length(pattern_array)-1
idx_final =
     1     2     3     4     5     6     9    10    11    12    13    14

Now find the indices which are not present the the idx_final indices

>> find(ismember(1:length(T.Var1),idx_final)==0)
%.....................^ Rows number in the table
ans =
     7     8

Third Option: normxcorr(): I have not tested with the sample given data

Fourth Way: There are multiple ways, choose as per your requirements and understanding

Once you get the rows deletion is easy

table_variable(idx,:)=[]

Hope it Helps :)

[Spencer Ferris]

Yes this worked perfectly! Thank you so much!