What is causing the reshaping error?

9 visualizaciones (últimos 30 días)
Calvin Chang
Calvin Chang el 8 de Jul. de 2019
Editada: Adam Danz el 11 de Jul. de 2019
I was playing with the number n, and sometime reshape gives an error when it is too small? Can somone help me deduce what is actually going on here?
M = 64;
k = log2(M);
n = 128;
numSamplesPerSymbol = 1;
rng default
dataIn = randi([0 1], n, 1);
stem(dataIn(1:128), 'filled');
title('Random Bits');
xlabel('Bit Index');
ylabel('Binary Value');
dataInMatrix = reshape(dataIn, length(dataIn)/k, k);
dataSymbolsIn = bi2de(dataInMatrix);
% dataSymbolsIn = b2d(dataInMatrix);
figure;
stem(dataSymbolsIn(1:50));
title('RandomYmbols');
xlabel('Symbol Index');
ylabel('Integer Value');

Respuesta aceptada

Adam Danz
Adam Danz el 9 de Jul. de 2019
Editada: Adam Danz el 11 de Jul. de 2019
As described in the documentation, reshape() the 2nd and 3rd inputs should be integers and when those two integers are multiplied, it should equal the number of elements in the first input. That makes sense since the 2nd and 3rd inputs define the dimensions of a matrix.
In the case of your example, length(dataIn)/k equals 21.333 so you're asking for a matrix with 21.3 rows which doesn't make sense.
For your script to work, n must be a multiple of k and k must be an integer.
But there are additional areas of concern. For example, log2() is used to define k and that may not return an interger. Secondly, in the line that calls stem(), you hard-coded 1:128 so if you use n=126, you'll get an indexing error at that line. Likewise, in the second call to stem(), you've hard-coded 1:50 which would cause another index error.
Instead of hard coding these indices, you 1:numel(...) or 1:length(...).

Más respuestas (0)

Categorías

Más información sobre Language Fundamentals 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!

Translated by