Reading in a text file with numbers and strings in

I have a .txt which I can read into a matrix fine.

I currently do this as follows:

FILENAME = 'inputData.csv';
sondeData = readmatrix(FILENAME,'numheaderlines',1); %ignore first line as this is heading

The text file has mostly numerical values. Along with some parameters which are presented as strings.

In those catagories the data is "GREEN" or "RED" which are essentially pass / fail status/

A sample of the data is below

inputData =

GREEN GREEN 11.88 876 1332 GREEN 758 GREEN 4.006 19.08 5.09265 GREEN

GREEN GREEN 12.09 863 1519 GREEN 761 GREEN 4.006 19.08 5.070016 GREEN

When I read the file in I get NaNs in all the columns that contain the RED / GREEN.

What I want to do is clean the data up - so remove any rows where the value in a certain column is RED. This means something has failed and so data is junk.

How can I load the data in such that I can do something like the below, to remove the junk rows

inputData(inputData(:, 2)== "RED", :)= []

I am not able to change how the inputData is provided to me - but I could do an intemediary step to turn those into booleans along the lines of GREEN = 1, RED = 0 if I had to.

Answer 1

[Star Strider]

It would help to haved the file.

The readtable function would likely be best for this, since it will also import the variable names (likely the first line of the file).

T1 = cell2table({'GREEN'  'GREEN'  11.88  876  1332  'GREEN'  758  'GREEN'  4.006  19.08  5.09265  'GREEN';
    'RED' 'RED' rand rand rand 'RED' rand 'RED' rand rand rand 'RED'; 
  'GREEN'  'GREEN'  12.09  863  1519  'GREEN'  761  'GREEN'  4.006  19.08  5.070016  'GREEN'})

T1 = 3×12 table
      Var1         Var2        Var3       Var4      Var5        Var6        Var7        Var8        Var9       Var10      Var11       Var12  
    _________    _________    _______    ______    _______    _________    _______    _________    _______    _______    _______    _________

    {'GREEN'}    {'GREEN'}      11.88       876       1332    {'GREEN'}        758    {'GREEN'}      4.006      19.08     5.0926    {'GREEN'}
    {'RED'  }    {'RED'  }    0.59669    0.9725    0.41824    {'RED'  }    0.18383    {'RED'  }    0.48095    0.27379    0.99147    {'RED'  }
    {'GREEN'}    {'GREEN'}      12.09       863       1519    {'GREEN'}        761    {'GREEN'}      4.006      19.08       5.07    {'GREEN'}

Lv = any(strcmpi(T1{:,[1 2 6 8 12]},'RED'),2)                                   % Test Variables

Lv = 3×1 logical array
   0
   1
   0

T1 = T1(~Lv,:)

T1 = 2×12 table
      Var1         Var2       Var3     Var4    Var5      Var6       Var7      Var8       Var9     Var10    Var11       Var12  
    _________    _________    _____    ____    ____    _________    ____    _________    _____    _____    ______    _________

    {'GREEN'}    {'GREEN'}    11.88    876     1332    {'GREEN'}    758     {'GREEN'}    4.006    19.08    5.0926    {'GREEN'}
    {'GREEN'}    {'GREEN'}    12.09    863     1519    {'GREEN'}    761     {'GREEN'}    4.006    19.08      5.07    {'GREEN'}

Experiment with the actual data.

EDIT — (18 Oct 2022 at 16:26)

Use curly braces for the table reference here —

a = inputData{1,3};

b = a^2;

.

Answer 2

[Les Beckham]

Try using readtable instead of readmatrix. This will read in the strings properly and allow you to do your cleanup based on the values of those strings.

[Richard Nash]

I looked at that - but when I tried it I couldn't seem to then do mathematical functions using the numerical data.

So if I read the example data in as a table and do the following:

I get an error "Operator '.^' is not supported for operands of type 'table'."

a = inputData(1,3);
b = a^2;