How to use fillmissing with table variables of type cell
2 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Hello,
I simply want to replace all empty cells ([]) by '99' in the table attached but only for the variables that begin with Q_. I get the following error however: Table variables of type cell are not supported.How do I get around that ?
all_vars = T1.Properties.VariableNames;
index = find(strncmp(all_vars, 'Q_', 2));
T1(:, index) = fillmissing(T1(:, index), 'constant', '99')
Thank you,
0 comentarios
Respuestas (1)
Guillaume
el 29 de Oct. de 2019
Well, you're a funny one!
You've asked a very similar question. You were given two answers, one of which used fillmissing to do exactly what you're asking now, but you accepted the other answer.
Note that my answer used a simpler way of finding the data variables starting with Q_. For a start the find serves no purpose in the code you show. You'd get the exact same result using
index = strncmp(all_vars, 'Q_', 2); %no need for find, you can use the logical vector directly
T(:, index) = ...
but using startsWith as per my answer is even simpler than strncmp.
2 comentarios
Guillaume
el 29 de Oct. de 2019
Which version of matlab are you using?
I don't get any error using R2019b:
>> a = {'A'; 'B'; 'C'};
o = {'A'; '[]'; 'C'};
Q_y = {'1'; '[]'; '3'};
T1 = table(Q_x, a, o, Q_y)
T1 =
3×4 table
Q_x a o Q_y
______ _____ ______ ______
{'1' } {'A'} {'A' } {'1' }
{'[]'} {'B'} {'[]'} {'[]'}
{'3' } {'C'} {'C' } {'3' }
>> missinglocs = ismissing(T1, '[]');
newt = fillmissing(T1, 'constant', '-99', 'DataVariables', startsWith(T1.Properties.VariableNames, 'Q_'), 'MissingLocations', missinglocs)
newt =
3×4 table
Q_x a o Q_y
_______ _____ ______ _______
{'1' } {'A'} {'A' } {'1' }
{'-99'} {'B'} {'[]'} {'-99'}
{'3' } {'C'} {'C' } {'3' }
Ver también
Categorías
Más información sobre Logical en Help Center y File Exchange.
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!