A Vector is a struct that holds information about a vector that has been allocated inside R. The data is:
long rows;
RData data;
double * ptr;You can use assignment and slicing as in languages like R and Matlab. Keep in mind that slices do not include the last element. Here’s an example:
// Create a new vector with space for five elements
auto x = Vector(5);
// Set all five elements to 1.7
x = 1.7;
// Print x with a message preceding it
x.print("Test vector");
// You can also use the empty slice operator to set all elements to the same scalar
x[] = -6.2;
x.print("Should be -6.2");
// You can use the empty slice operator to copy in the elements of a double[]
// There is a check that the lengths match
x[] = [1.1, 2.2, 3.3, 4.4, 5.5];
x.print("Should be increasing values");
// You can set a subset of the vector to the same scalar
x[2..5] = -1.2;
x.print("Last three elements should be -1.2");
// You can also set a subset of the vector to the elements of a double[]
// There's a check that the lengths match
x[2..5] = [-0.2, -0.3, -0.4];
x.print("Last three elements changed");You can pass a string of R code and store the result as a Vector.
auto x = Vector(rnorm(10));You can send the number of elements and a Vector with that many elements will be allocated. The elements are initialized to zero.
auto x = Vector(12)This is used for assigning to a zero-length Vector. Otherwise there is no way to assign to a Vector that is part of a struct that has already been constructed, since the bounds checking will fail.
Makes a copy in R, but forces the allocation of a new Vector.
auto y3 = Vector(x);Makes a copy in R, but forces the allocation of a new Vector.
Allocates a new Vector and copies the elements of v into it.
auto x = Vector([1.1, 2.2, 3.3]);// Indexing
v[1]
// Grab the first, third, and fifth elements
v[[0, 2, 4]]
// Set elements
v[1] = 3.3;Get one element. Indexing starts at zero.
Return a new vector holding the values at the indexes in obs.
Assign to the element at index r.
Checks that the number of elements match. The exception is if there are zero rows, which is taken to mean this has not been initialized. Will allocate the vector and copy the elements of v into it.
Checks that the number of elements match. The exception is if there are zero rows, which is taken to mean this has not been initialized. Will allocate the vector and copy the elements of v into it.
this[0..3] returns a new Vector with three elements. j is not included.
this[] returns a new Vector with all elements of this.
This does the same thing as opAssign.
This does the same thing as opAssign.
A Vector is a range. You can do this type of thing:
auto v = Vector([1.1, 2.2, 3.3]);
foreach(val; Vector([1.1, 2.2, 3.3])) {
writeln(val);
}
import std.range: enumerate;
foreach(ii, val; Vector([1.1, 2.2, 3.3]).enumerate) {
writeln(ii+1, " ", val);
}The last value
Allows you to do v[3..$]
Returns a Matrix holding the data in this as one column.
Returns a Matrix holding the data in this as one row.
Convert this to a double[] using std.conv.to:
auto v = Vector([1.1, 2.2, 3.3]);
double[] vv = v.to!(double[]);Allows element-by-element operations like this:
auto v = Vector([1.1, 2.2, 3.3]);
auto v2 = Vector([4.4, 5.5, 6.6]);
v += v2;
v -= v2;
v *= v2;
v /= v2;Calls R’s head function. Returns a new Vector of length n. If n is negative, drops n elements from the end, so the length is v.length - n.
Calls R’s tail function. Returns a new Vector of length n. If n is negative, drops n elements from the front, so the length is v.length - n.
Used to fill the elements of a Vector. The main case is where you want to refill a Vector inside a loop and you’d like to avoid allocation for performance reasons. Fill is a reference to the Vector.
auto vraw = Vector(10);
auto v = Fill(vraw);
v ~= 1.1;
v ~= 2.2;
v ~= 3.3;
writeln(v.full()); // false, because the capacity is 7
writeln(v.capacity()); // 7
v ~= [4.4, 5.5, 6.6, 7.7, 8.8, 9.9, 11.0];
writeln(v.full()); // true, because the capacity is 0
v ~= 12.1; // Error, because there's no place to put another element
vraw.print("The vector"); // Work with vraw for anything other than filling
v.reset(); // Start over without creating garbage and avoiding allocationSame as Fill, but in case you only want to work with part of a Vector rather than the whole thing. The main case is where you’re only updating some of the elements inside a loop because you’re either not updating all elements or you’re doing the update in several places.
auto vraw = Vector(10);
auto v = PartialFill(vraw, 3, 7); // You only want to fill vraw[3]..vraw[6]Vector holds double precision values. IntVector holds integer values.
Vector holds double precision values. BoolVector holds true/false values. Under the covers, since we’re working with C, it holds int values. The interface is designed to feel like you’re working with bool values.
auto bv = BoolVector(2);
bv[0] = true;
bv[1] = false;
writeln(bv[0]); // true, even though ptr[0] is 1Create and work with a vector of R strings. Working with the underlying data is messy (they’re C strings, after all). You should work with a StringVector instead of working with the R data directly.
auto sv = StringVector(["Foo", "Bar", "Baz", "D is kind of cool!"]);
sv.print("A string vector");
writeln(sv[1]);