use runmat_accelerate_api::GpuTensorHandle;
use runmat_builtins::{CharArray, ComplexTensor, Tensor, Value};
use runmat_macros::runtime_builtin;
use crate::builtins::common::spec::{
BroadcastSemantics, BuiltinFusionSpec, BuiltinGpuSpec, ConstantStrategy, FusionError,
FusionExprContext, FusionKernelTemplate, GpuOpKind, ProviderHook, ReductionNaN,
ResidencyPolicy, ScalarType, ShapeRequirements,
};
use crate::builtins::common::{gpu_helpers, map_control_flow_with_builtin, tensor};
use crate::builtins::math::type_resolvers::numeric_unary_type;
use crate::{build_runtime_error, BuiltinResult, RuntimeError};
#[runmat_macros::register_gpu_spec(builtin_path = "crate::builtins::math::elementwise::conj")]
pub const GPU_SPEC: BuiltinGpuSpec = BuiltinGpuSpec {
name: "conj",
op_kind: GpuOpKind::Elementwise,
supported_precisions: &[ScalarType::F32, ScalarType::F64],
broadcast: BroadcastSemantics::Matlab,
provider_hooks: &[ProviderHook::Unary { name: "unary_conj" }],
constant_strategy: ConstantStrategy::InlineLiteral,
residency: ResidencyPolicy::NewHandle,
nan_mode: ReductionNaN::Include,
two_pass_threshold: None,
workgroup_size: None,
accepts_nan_mode: false,
notes:
"Providers may execute conj in-place for real tensors via unary_conj; complex tensors currently gather to the host for conjugation.",
};
#[runmat_macros::register_fusion_spec(builtin_path = "crate::builtins::math::elementwise::conj")]
pub const FUSION_SPEC: BuiltinFusionSpec = BuiltinFusionSpec {
name: "conj",
shape: ShapeRequirements::BroadcastCompatible,
constant_strategy: ConstantStrategy::InlineLiteral,
elementwise: Some(FusionKernelTemplate {
scalar_precisions: &[ScalarType::F32, ScalarType::F64],
wgsl_body: |ctx: &FusionExprContext| {
let input = ctx
.inputs
.first()
.ok_or(FusionError::MissingInput(0))?;
Ok(format!("({input})"))
},
}),
reduction: None,
emits_nan: false,
notes:
"Fusion kernels treat conj as an identity for real tensors; complex tensors fall back to the CPU path until native complex fusion is available.",
};
const BUILTIN_NAME: &str = "conj";
fn builtin_error(message: impl Into<String>) -> RuntimeError {
build_runtime_error(message)
.with_builtin(BUILTIN_NAME)
.build()
}
#[runtime_builtin(
name = "conj",
category = "math/elementwise",
summary = "Compute the complex conjugate of scalars, vectors, matrices, or N-D tensors.",
keywords = "conj,complex conjugate,complex,elementwise,gpu",
accel = "unary",
type_resolver(numeric_unary_type),
builtin_path = "crate::builtins::math::elementwise::conj"
)]
async fn conj_builtin(value: Value) -> BuiltinResult<Value> {
match value {
Value::GpuTensor(handle) => conj_gpu(handle).await,
Value::Complex(re, im) => conj_complex_scalar(re, im),
Value::ComplexTensor(ct) => conj_complex_tensor(ct),
Value::CharArray(ca) => conj_char_array(ca),
Value::String(_) | Value::StringArray(_) => {
Err(builtin_error("conj: expected numeric input"))
}
x @ (Value::Tensor(_)
| Value::LogicalArray(_)
| Value::Num(_)
| Value::Int(_)
| Value::Bool(_)) => conj_real(x),
other => Err(builtin_error(format!(
"conj: unsupported input type {:?}; expected numeric, logical, or char data",
other
))),
}
}
async fn conj_gpu(handle: GpuTensorHandle) -> BuiltinResult<Value> {
if let Some(provider) = runmat_accelerate_api::provider_for_handle(&handle) {
if let Ok(out) = provider.unary_conj(&handle).await {
return Ok(Value::GpuTensor(out));
}
}
let tensor = gpu_helpers::gather_tensor_async(&handle)
.await
.map_err(|flow| map_control_flow_with_builtin(flow, BUILTIN_NAME))?;
Ok(tensor::tensor_into_value(conj_tensor(tensor)?))
}
fn conj_real(value: Value) -> BuiltinResult<Value> {
let tensor = tensor::value_into_tensor_for("conj", value)
.map_err(|e| builtin_error(format!("conj: {e}")))?;
Ok(tensor::tensor_into_value(conj_tensor(tensor)?))
}
fn conj_tensor(tensor: Tensor) -> BuiltinResult<Tensor> {
Ok(tensor)
}
fn conj_complex_scalar(re: f64, im: f64) -> BuiltinResult<Value> {
let imag = -im;
if imag == 0.0 && !imag.is_nan() {
Ok(Value::Num(re))
} else {
Ok(Value::Complex(re, imag))
}
}
fn conj_complex_tensor(ct: ComplexTensor) -> BuiltinResult<Value> {
let ComplexTensor {
data: ct_data,
shape,
..
} = ct;
let mut all_real = true;
let mut data = Vec::with_capacity(ct_data.len());
for (re, im) in ct_data {
let imag = -im;
if imag != 0.0 || imag.is_nan() {
all_real = false;
}
data.push((re, imag));
}
if all_real {
let real: Vec<f64> = data.into_iter().map(|(re, _)| re).collect();
let tensor =
Tensor::new(real, shape.clone()).map_err(|e| builtin_error(format!("conj: {e}")))?;
Ok(tensor::tensor_into_value(tensor))
} else {
let tensor =
ComplexTensor::new(data, shape).map_err(|e| builtin_error(format!("conj: {e}")))?;
Ok(Value::ComplexTensor(tensor))
}
}
fn conj_char_array(ca: CharArray) -> BuiltinResult<Value> {
let data = ca
.data
.iter()
.map(|&ch| ch as u32 as f64)
.collect::<Vec<_>>();
let tensor = Tensor::new(data, vec![ca.rows, ca.cols])
.map_err(|e| builtin_error(format!("conj: {e}")))?;
Ok(tensor::tensor_into_value(tensor))
}
#[cfg(test)]
pub(crate) mod tests {
use super::*;
use crate::builtins::common::test_support;
use futures::executor::block_on;
use runmat_builtins::{IntValue, LogicalArray, ResolveContext, Type};
fn conj_builtin(value: Value) -> BuiltinResult<Value> {
block_on(super::conj_builtin(value))
}
#[test]
fn conj_type_preserves_tensor_shape() {
let out = numeric_unary_type(
&[Type::Tensor {
shape: Some(vec![Some(2), Some(3)]),
}],
&ResolveContext::new(Vec::new()),
);
assert_eq!(
out,
Type::Tensor {
shape: Some(vec![Some(2), Some(3)])
}
);
}
#[test]
fn conj_type_scalar_tensor_returns_num() {
let out = numeric_unary_type(
&[Type::Tensor {
shape: Some(vec![Some(1), Some(1)]),
}],
&ResolveContext::new(Vec::new()),
);
assert_eq!(out, Type::Num);
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_scalar_real() {
let result = conj_builtin(Value::Num(-2.5)).expect("conj");
match result {
Value::Num(n) => assert!((n + 2.5).abs() < 1e-12),
other => panic!("expected scalar result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_complex_scalar() {
let result = conj_builtin(Value::Complex(3.0, 4.0)).expect("conj");
match result {
Value::Complex(re, im) => {
assert!((re - 3.0).abs() < 1e-12);
assert!((im + 4.0).abs() < 1e-12);
}
other => panic!("expected complex result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_complex_scalar_zero_imag_returns_real() {
let result = conj_builtin(Value::Complex(5.0, 0.0)).expect("conj");
match result {
Value::Num(n) => assert!((n - 5.0).abs() < 1e-12),
other => panic!("expected scalar result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_promotes_logical_to_double() {
let logical =
LogicalArray::new(vec![0, 1, 1, 0], vec![2, 2]).expect("logical array construction");
let result = conj_builtin(Value::LogicalArray(logical)).expect("conj");
match result {
Value::Tensor(t) => {
assert_eq!(t.shape, vec![2, 2]);
assert_eq!(t.data, vec![0.0, 1.0, 1.0, 0.0]);
}
other => panic!("expected tensor result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_int_promotes_to_double() {
let result = conj_builtin(Value::Int(IntValue::I32(7))).expect("conj");
match result {
Value::Num(n) => assert!((n - 7.0).abs() < 1e-12),
other => panic!("expected scalar result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_complex_tensor_to_complex_tensor() {
let tensor =
ComplexTensor::new(vec![(1.0, 2.0), (-3.0, -4.0)], vec![2, 1]).expect("complex tensor");
let result = conj_builtin(Value::ComplexTensor(tensor)).expect("conj");
match result {
Value::ComplexTensor(ct) => {
assert_eq!(ct.shape, vec![2, 1]);
assert_eq!(ct.data[0], (1.0, -2.0));
assert_eq!(ct.data[1], (-3.0, 4.0));
}
other => panic!("expected complex tensor, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_complex_tensor_realises_real_when_imag_zero() {
let tensor =
ComplexTensor::new(vec![(1.0, 0.0), (2.0, -0.0)], vec![2, 1]).expect("complex tensor");
let result = conj_builtin(Value::ComplexTensor(tensor)).expect("conj");
match result {
Value::Tensor(t) => {
assert_eq!(t.shape, vec![2, 1]);
assert_eq!(t.data, vec![1.0, 2.0]);
}
other => panic!("expected real tensor, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_char_array_returns_double_codes() {
let chars = CharArray::new("Hi".chars().collect(), 1, 2).expect("char array");
let result = conj_builtin(Value::CharArray(chars)).expect("conj");
match result {
Value::Tensor(t) => {
assert_eq!(t.shape, vec![1, 2]);
assert_eq!(t.data, vec![72.0, 105.0]);
}
other => panic!("expected tensor result, got {other:?}"),
}
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_errors_on_string_input() {
let err = conj_builtin(Value::from("hello")).unwrap_err();
assert!(err.message().contains("expected numeric input"));
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
fn conj_gpu_provider_roundtrip() {
test_support::with_test_provider(|provider| {
let tensor = Tensor::new(vec![0.0, 1.0, -3.0, 4.0], vec![4, 1]).unwrap();
let view = runmat_accelerate_api::HostTensorView {
data: &tensor.data,
shape: &tensor.shape,
};
let handle = provider.upload(&view).expect("upload");
let result = conj_builtin(Value::GpuTensor(handle)).expect("conj");
let gathered = test_support::gather(result).expect("gather");
assert_eq!(gathered.shape, vec![4, 1]);
assert_eq!(gathered.data, tensor.data);
});
}
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
#[test]
#[cfg(feature = "wgpu")]
fn conj_wgpu_matches_cpu_for_real() {
let _ = runmat_accelerate::backend::wgpu::provider::register_wgpu_provider(
runmat_accelerate::backend::wgpu::provider::WgpuProviderOptions::default(),
);
let tensor = Tensor::new(vec![1.0, -2.0, 3.5, 0.0], vec![4, 1]).unwrap();
let cpu = conj_real(Value::Tensor(tensor.clone())).unwrap();
let view = runmat_accelerate_api::HostTensorView {
data: &tensor.data,
shape: &tensor.shape,
};
let handle = runmat_accelerate_api::provider()
.unwrap()
.upload(&view)
.unwrap();
let gpu = block_on(conj_gpu(handle)).unwrap();
let gathered = test_support::gather(gpu).expect("gather");
match (cpu, gathered) {
(Value::Tensor(ct), gt) => {
assert_eq!(ct.shape, gt.shape);
assert_eq!(ct.data, gt.data);
}
_ => panic!("unexpected shapes"),
}
}
}