What is the difference between matrix() and as.matrix() in r?

Question

I ran the following in R and received the same output for both matrix() and as.matrix() and now I am not sure what the difference between them is:

Answer 1

matrix constructs a matrix from its first argument, with a given number of rows and columns. If the supplied object isn't large enough for the desired output, matrix will recycle its elements: for example, matrix(1:2), nrow=3, ncol=4). Conversely, if the object is too big, then the surplus elements will be dropped: for example, matrix(1:20, nrow=3, ncol=4).

as.matrix converts its first argument into a matrix, the dimensions of which will be inferred from the input.

Answer 2

matrix creates a matrix from the given set of values. as.matrix attempts to turn its argument into a matrix.

Further, matrix() makes efforts to keep logical matrices logical, i.e., and to determine specially structured matrices such as symmetric, triangular or diagonal ones.

as.matrix is a generic function. The method for data frames will return a character matrix if there is only atomic columns and any non-(numeric/logical/complex) column, applying as.vector to factors and format to other non-character columns. Otherwise, the usual coercion hierarchy (logical < integer < double < complex) will be used, e.g., all-logical data frames will be coerced to a logical matrix, mixed logical-integer will give a integer matrix, etc.

The default method for as.matrix calls as.vector(x), and hence e.g. coerces factors to character vectors.

Answer 3

matrix takes data and further arguments nrow and ncol.

?matrix
 If one of ‘nrow’ or ‘ncol’ is not given, an attempt is made to
 infer it from the length of ‘data’ and the other parameter.  If
 neither is given, a one-column matrix is returned.

as.matrix is a method with different behaviours for different types, but mainly to give back an n*m matrix from an n*m input.

?as.matrix
 ‘as.matrix’ is a generic function.  The method for data frames
 will return a character matrix if there is only atomic columns and
 any non-(numeric/logical/complex) column, applying ‘as.vector’ to
 factors and ‘format’ to other non-character columns.  Otherwise,
 the usual coercion hierarchy (logical < integer < double <
 complex) will be used, e.g., all-logical data frames will be
 coerced to a logical matrix, mixed logical-integer will give a
 integer matrix, etc.

The difference between them comes primarily from the shape of the input, matrix doesn't care about the shape, as.matrix does and will maintain it (though the details depend on the actual methods for the input, and in your case a dimensionless vector corresponds to a single column matrix.) It doesn't matter if the input is raw, logical, integer, numeric, character, or complex, etc.