Create function to convert data type as table: 2 errors

Hello, I can run the code in a *.mlx file and its great but the code does not work as a function. Please help.

*.mlx file

h = d.Vertical_Steering_Command  % selected a specific variable, categorial (Atltitude, Limbo)

*.m file (function call)

function Vertical_Steering_Cmd_num = data_conversion(h)

% convert Vertical Steering Command categorical data to double data type as Table 'h'
% note: Altitude is set to '0', Limbo is set to '6' 

h = zeros(size(h.Vertical_Steering_Command));  % error 
h(ismember(h.Vertical_Steering_Command,'Limbo'))=6;

% Output double data: add double data type from table 'h' and 
% add new column for Vertical Steering Command number data to table 'h' 
h.Vertical_Steering_Cmd_num = h
end

------------------------------

Error using indexing

Attempt to reference field of non-structure array.

Error in data_converstion (line 7)

h = zeros(size(h.Vertical_Steering_Command));

Answer 1

[Voss]

One problem appears to be that the function expects a table but you're giving it a categorical array. I can't say for sure because I don't know how you're calling the function, but if you're using the h you define as

h = d.Vertical_Steering_Command  % selected a specific variable, categorial (Atltitude, Limbo)

then that's a categorical array (if your comment is accurate).

Another problem is that you're using the variable h to mean two different things inside the function: the input table and the numeric data you want to add to the table.

Try defining the function as follows. I changed the numeric data variable to data (I'm not sure what the output should be so I output the new table h, with the added column).

function h = data_conversion(h)

% convert Vertical Steering Command categorical data to double data type as Table 'h'
% note: Altitude is set to '0', Limbo is set to '6' 

data = zeros(size(h.Vertical_Steering_Command));  % error 
data(ismember(h.Vertical_Steering_Command,'Limbo'))=6;

% Output double data: add double data type from table 'h' and 
% add new column for Vertical Steering Command number data to table 'h' 
h.Vertical_Steering_Cmd_num = data;
end

And when you call it, give it a table:

d = data_conversion(d); % d is your table; the function updates the table and returns it, and the new table is stored as d again

[Deborah Johnson]

Hello Voss,

The program is still not running. Should I convert the array to table before passing it into the function?

[Deborah Johnson]

Voss, I also just selected 'Run' function, I am told it needs more inputs? Thank you for the information about the array. I orginally pulled the data from a table and ran it code in the *,mls code and it works great. But I am asked now to create a 'function' to handle the same process. When I run the new function, it does not output the same error until I run the 'd = data_conversion(h). Is there another method that would work?

Debbie

[Voss]

"Should I convert the array to table before passing it into the function?"

The comments in the function in your question suggest that h inside the function is a table, so I was going by that assumption.

"I also just selected 'Run' function, I am told it needs more inputs?"

The function data_conversion (both yours and mine) takes one input, so you have to supply it one input, which is done by running a command like:

d = data_conversion(d)

It won't run with zero inputs, which is what clicking the Run button does.

Here's a more complete code, starting with reading the xlsx file you attached:

d = readtable('Book1.xlsx') % note that this has 'Vertical_Steering' not 'Vertical_Steering_Command', so I modified the function below                         

d = 1045×1 table
    Vertical_Steering
    _________________

      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   
      {'Altitude'}   

d = data_conversion(d) % the new column is added

d = 1045×2 table
    Vertical_Steering    Vertical_Steering_Cmd_num
    _________________    _________________________

      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            
      {'Altitude'}                   0            

function h = data_conversion(h)

% convert Vertical Steering Command categorical data to double data type as Table 'h'
% note: Altitude is set to '0', Limbo is set to '6' 

data = zeros(size(h.Vertical_Steering));  % error 
data(ismember(h.Vertical_Steering,'Limbo'))=6;

% Output double data: add double data type from table 'h' and 
% add new column for Vertical Steering Command number data to table 'h' 
h.Vertical_Steering_Cmd_num = data;

end

[Deborah Johnson]

Hello Voss, I apologize for not getting back with you sooner but have been busy. Thank you about letting me know about the 'zeros'. The conversion is working now. Debbie

[Voss]

Debbie, I'm glad it's working!

[Deborah Johnson]

Voss, can I call up the readtable with Matlab instead of xlsx? h=readtable(d.Vertical_Steering) If not, what do you recommend?

[Voss]

readtable is for reading a file (e.g., a spreadsheet file such as an .xlsx file) and returning the contents as a table variable.

I'm not sure what you have in mind. Can you explain?

[Deborah Johnson]

Hello Voss, I have a cvs table with 20 columns of data. Only 2 columns have categorical data that require conversion to double data type with each as a standalone function. Each categorigal data column can have up to 12 data conversions ranging from 0 to 12. Due to the zeros call,

I'm looking for an alternate method. I could call the the original cvs file but it is really large. What do you suggest?

[Voss]

Do you get an "Out Of Memory" error when using zeros?

[Deborah Johnson]

No. The orgiinal variable names are quite long when the original *.cvs file is read as readtable. I am told to set Variable Nmes to preserve.

I have experimented today trying different methods. Such as renaming that one particular variable name within the function but that didn't work. Then tried reading the MATLAB table as ' h = table(h.Vertical_Steering_Cmd), h = table(h,"Vertical_Steering_Cmd).

The function and matlab code is below with new attached file. Any suggestions?

function h = data_conversion(h)
% convert Vertical Steering Command categorical data to double data type as Table 'h'
% note: Altitude is set to '0' 
h = readtable(h, 'Vertical_Steering_Cmd');
data = zeros(size(h.Vertical_Steering_Cmd));
data(ismember(h.Vertical_Steering_Cmd,'Limbo')) = 6;
% copy number data from table 'h' and 
% add new column for Vertical Steering number data to table 'h' 
h.Vertical_Steering_Cmd_num = data;
end

---------------------------

MATLABf

f = readtable("two_var_columns.csv");
% rename all variable headers in sequential order
d = renamevars(d,["FLTSIMocAutofltseqLateral_Steering_Command", 
    "FLTSIMocAutofltseqVertical_Steering_Command"],... 
    ["Lateral_Steering_Cmd", "Vertical_Steering_Cmd"]);
% delete 1st row of erowgenios data
d([1],:) = [];
h = d.Vertical_Steering_Cmd
h = data_conversion(h.Vertical_Steering_Cnd)

error using indexing, Attempt to reference field of a non-structure array

[Walter Roberson]

h = d.Vertical_Steering_Cmd

d is a table. d.Vertical_Steering_Cmd is the content of the variable Vertical_Steering_Cmd within the table.

h = data_conversion(h.Vertical_Steering_Cnd)

h is not a table or a struct or an object. You already extracted Vertical_Steering_Cmd into h so h.Vertical_Steering_Cnd would be like trying to use d.Vertical_Steering_Cmd.Vertical_Steering_Cnd -- not going to work.

[Deborah Johnson]

Hello Voss and Walter, Thank you for assisting and your support. I apologize for my lack of experience. My co-worker and I figured out that I was trying to be too correct because I was specifying a table.variable_name which was actually an "array.variable_column" When I took off the variable_column, the function worked correctly and completes all of the data type conversions!

The function's solution:

----------------------

function i = data_conversion(h)
% convert Vertical Steering Command categorical data to double data type as Table 'h'
% note: Altitude is set to '0'  
data = zeros(size(h));
data(ismember(h(:),'Limbo')) = 6;
i = data;
end

-----------------------

Thank you again!