Data types

EN: Stage 5 Value::Sparse stores rows, cols and sorted unique finite nonzero triplets; class is double, issparse is true. Logical rows*cols<=16777216 and input/stored triplets<=1000000, not unbounded sparse storage. See /language/sparse-advanced-solvers for sparse/full/issparse/nnz and solver boundaries. Studio read-only sparse metadata is upcoming, not an editable grid.

PT: Value::Sparse da etapa 5 guarda rows, cols e tripletas ordenadas unicas finitas nao zero; classe double, issparse true. Produto logico rows*cols<=16777216 e tripletas de entrada/armazenadas<=1000000, nao armazenamento ilimitado. Veja /language/sparse-advanced-solvers para sparse/full/issparse/nnz e limites dos solvers. Metadados esparsos somente leitura no Studio estao previstos, nao grade editavel.

Sun stores real and complex scalars/arrays as double-precision values. Logical values have dense 2D array storage. Strings remain scalar, and runtime exceptions use the SunException class. Stage 4 adds scalar structs, 2D cells, string-key maps, tables, explicit null and UTC datetime scalars; see /language/data-and-files for the bounded subset.

• double: real scalars/arrays and complex scalars/arrays
• logical: boolean scalars and dense logical arrays
• string: UTF-8 text scalars (char converts to the same type)
• struct, cell, containers.Map, table: value-copy containers
• null: explicit null, not []
• datetime: canonical UTC scalar
• SunException: caught runtime exceptions

class returns the exact runtime class. isa accepts exact names and the numeric categories numeric, float, and integer.

disp(class(3.5));              % double
disp(class(true));             % logical
disp(isa([1, 2], "numeric")); % true
disp(isa(1, "integer"));     % false

double preserves double arrays and converts logical scalars/arrays to 0 or 1. real, imag, and conj preserve complex-array shapes. char and string convert scalar text, finite numbers, logicals, datetime and null to scalar Unicode text; they do not introduce MATLAB character arrays or string arrays.

mask = [true, false];
disp(double(mask));       % [1, 0]
z = [1+i, 2-i];
disp(real(z));
disp(imag(z));

• Advanced signal processing and mature complex linear algebra beyond FFT/eig/roots; see /language/spectral-statistics
• Typed dense arrays for single, half, and signed or unsigned integers
• Separate character arrays and string arrays
• Datetime arrays, arithmetic, durations and categorical arrays
• Struct arrays, timetables, joins and OOP beyond the Stage 4 scalar/container subset
• Time series and broader map APIs