How to change strange cell array within a table to double or categorical array

I have some columns within a table that are labeled as cells that contain both NaN's and "Checked" (a categorical).

Please see the attach screen shot.

How to I change those columns to either a categorical array or to a numeric array with "checked" as 1's?

Thanks!

Answer 1

[dpb]

Well, it'd be easier to be sure if could actually know for sure what the data storage is, but guessing on account of the " surrounding the string data, I created a dummy cell array as

LS=num2cell(nan(10,1));                 % start w/ NaN
LS(randperm(10,3))={"Checked"};         % set some random cells to "Checked"

% Result for this invocation was
>> LS
LS =
  10×1 cell array
    {[      NaN]}
    {[      NaN]}
    {["Checked"]}
    {["Checked"]}
    {[      NaN]}
    {[      NaN]}
    {[      NaN]}
    {[      NaN]}
    {[      NaN]}
    {["Checked"]}
>>

To convert to categorical is relatively straightforward; the NaN

ix=~cellfun(@isstring,LS);      % Locate the NaN (isnan fails on string cells so complement)
LS(ix)={"Not Checked"};         % Put another string there--your choice as to what...
categorical(cellstr(LS))        % categorical uses unique() which can't stomach string array input
ans = 
  10×1 categorical array
     Not Checked 
     Not Checked 
     Checked 
     Checked 
     Not Checked 
     Not Checked 
     Not Checked 
     Not Checked 
     Not Checked 
     Checked 
>> 

Alternatively, the original idea would have been something like

ix=cellfun(@isstring,LS);           % this time we really _do_ want the strings
LS(ix)={1};                         % turn them to 1s for now to be all numeric array
% This leaves us with
>> LS
LS =
  10×1 cell array
    {[NaN]}
    {[NaN]}
    {[  1]}
    {[  1]}
    {[NaN]}
    {[NaN]}
    {[NaN]}
    {[NaN]}
    {[NaN]}
    {[  1]}
>> 

% convert to categorical, for NaN to be in allowable dataset, define outputs
>> categorical(cell2mat(LS),[nan 1],["UnChecked";"Checked"])
ans = 
  10×1 categorical array
     UnChecked 
     UnChecked 
     Checked 
     Checked 
     UnChecked 
     UnChecked 
     UnChecked 
     UnChecked 
     UnChecked 
     Checked 
>> 

It's pretty much just your choice which way to proceed...

Answer 2

[dpb]

Use the optional valueset, and perhaps catnames inputs to categorical to define as desired.

[Flynn McGuire]

This doesn't seem to work becuase of the NaN's mixed in the column