diff --git a/b_asic/scheduler.py b/b_asic/scheduler.py
index 306cdccd5a5d492434459cce925cc987c5bee96b..41ed7044831049f4edb59296de443d54363e06d7 100644
--- a/b_asic/scheduler.py
+++ b/b_asic/scheduler.py
@@ -1020,6 +1020,8 @@ class RecursiveListScheduler(ListScheduler):
destination_port = output_port.signals[0].destination
destination_op = destination_port.operation
if destination_op.graph_id not in self._remaining_ops:
+ if isinstance(destination_op, Delay):
+ continue
# spotted a recursive operation -> check if ok
op_available_time = (
self._current_time + op.latency_offsets[f"out{output_port.index}"]
diff --git a/b_asic/sfg_generators.py b/b_asic/sfg_generators.py
index faf79d2b6175f70646e81b4289cb9356bd4db939..7c85ad56188be85f527cecf5c25ebc2042c00275 100644
--- a/b_asic/sfg_generators.py
+++ b/b_asic/sfg_generators.py
@@ -164,7 +164,7 @@ def direct_form_fir(
See Also
--------
- transposed_direct_form_fir
+ transposed_direct_form_fir, symmetric_fir
"""
np_coefficients = np.atleast_1d(np.squeeze(np.asarray(coefficients)))
taps = len(np_coefficients)
@@ -232,7 +232,7 @@ def transposed_direct_form_fir(
See Also
--------
- direct_form_fir
+ direct_form_fir, symmetric_fir
"""
np_coefficients = np.atleast_1d(np.squeeze(np.asarray(coefficients)))
taps = len(np_coefficients)
@@ -265,6 +265,85 @@ def transposed_direct_form_fir(
return SFG([input_op], [output], name=Name(name))
+def symmetric_fir(
+ coefficients: Sequence[complex],
+ name: str | None = None,
+ mult_properties: dict[str, int] | dict[str, dict[str, int]] | None = None,
+ add_properties: dict[str, int] | dict[str, dict[str, int]] | None = None,
+) -> SFG:
+ r"""Generate a signal flow graph of a symmetric FIR filter.
+
+ The *coefficients* parameter is a sequence of impulse response values of even length::
+
+ coefficients = [h0, h1, h2, ..., hN]
+
+ Leading to the transfer function:
+
+ .. math:: \sum_{i=0}^N h_iz^{-i}
+
+ Parameters
+ ----------
+ coefficients : 1D-array
+ Coefficients to use for the FIR filter section.
+ name : Name, optional
+ The name of the SFG. If None, "Transposed direct-form FIR filter".
+ mult_properties : dictionary, optional
+ Properties passed to :class:`~b_asic.core_operations.ConstantMultiplication`.
+ add_properties : dictionary, optional
+ Properties passed to :class:`~b_asic.core_operations.Addition`.
+
+ Returns
+ -------
+ Signal flow graph
+
+ See Also
+ --------
+ direct_form_fir, transposed_direct_form_fir
+ """
+ np_coefficients = np.atleast_1d(np.squeeze(np.asarray(coefficients)))
+ taps = len(np_coefficients)
+ if not taps:
+ raise ValueError("Coefficients cannot be empty")
+ if taps > 1 and taps % 2 != 0:
+ raise ValueError("Coefficients must be of even length")
+ if np_coefficients.ndim != 1:
+ raise TypeError("coefficients must be a 1D-array")
+ if name is None:
+ name = "Symmetric FIR filter"
+ if mult_properties is None:
+ mult_properties = {}
+ if add_properties is None:
+ add_properties = {}
+ input_op = Input()
+ output = Output()
+
+ delays = [input_op]
+ for _ in range(taps - 1):
+ delays.append(Delay(delays[-1]))
+
+ add_layer_1 = [
+ Addition(delays[i], delays[-i - 1], **add_properties) for i in range(taps // 2)
+ ]
+
+ if taps == 1:
+ muls = [ConstantMultiplication(coefficients[0], input_op, **mult_properties)]
+ else:
+ muls = [
+ ConstantMultiplication(coefficients[i], add_layer_1[i], **mult_properties)
+ for i in range(taps // 2)
+ ]
+
+ previous_op = muls[0]
+ add_layer_2 = []
+ for i in range(taps // 2 - 1):
+ add_layer_2.append(Addition(previous_op, muls[i + 1], **add_properties))
+ previous_op = add_layer_2[-1]
+
+ output <<= add_layer_2[-1] if add_layer_2 else muls[0]
+
+ return SFG([input_op], [output], name=Name(name))
+
+
def direct_form_1_iir(
b: Sequence[complex],
a: Sequence[complex],
diff --git a/test/unit/test_sfg_generators.py b/test/unit/test_sfg_generators.py
index 569528b88c0875209d6208aa3b3fb14d3851e1d9..f164ee50866565da9b20d4c8e2e58ab5921641bb 100644
--- a/test/unit/test_sfg_generators.py
+++ b/test/unit/test_sfg_generators.py
@@ -16,6 +16,7 @@ from b_asic.sfg_generators import (
direct_form_fir,
ldlt_matrix_inverse,
radix_2_dif_fft,
+ symmetric_fir,
transposed_direct_form_fir,
wdf_allpass,
)
@@ -235,8 +236,103 @@ def test_transposed_direct_form_fir():
assert len([comp for comp in sfg.components if isinstance(comp, Delay)]) == 0
+def test_symmetric_fir():
+ impulse_response = [0.3, 0.5, 0.5, 0.3]
+ sfg = symmetric_fir(impulse_response)
+ assert (
+ len(
+ [
+ comp
+ for comp in sfg.components
+ if isinstance(comp, ConstantMultiplication)
+ ]
+ )
+ == 2
+ )
+ assert len([comp for comp in sfg.components if isinstance(comp, Addition)]) == 3
+ assert len([comp for comp in sfg.components if isinstance(comp, Delay)]) == 3
+
+ sim = Simulation(sfg, [Impulse()])
+ sim.run_for(5)
+ impulse_response.append(0.0)
+ assert np.allclose(sim.results["0"], impulse_response)
+
+ impulse_response = [0.1, 0.2, 0.3, 0.4, 0.4, 0.3, 0.2, 0.1]
+ sfg = symmetric_fir(
+ impulse_response,
+ mult_properties={"latency": 2, "execution_time": 1},
+ add_properties={"latency": 1, "execution_time": 1},
+ )
+ assert sfg.critical_path_time() == 6
+
+ sim = Simulation(sfg, [Impulse()])
+ sim.run_for(9)
+ impulse_response.append(0.0)
+ assert np.allclose(sim.results["0"], impulse_response)
+
+ impulse_response = [0.3]
+ sfg = symmetric_fir(impulse_response)
+ assert (
+ len(
+ [
+ comp
+ for comp in sfg.components
+ if isinstance(comp, ConstantMultiplication)
+ ]
+ )
+ == 1
+ )
+ assert len([comp for comp in sfg.components if isinstance(comp, Addition)]) == 0
+ assert len([comp for comp in sfg.components if isinstance(comp, Delay)]) == 0
+
+ impulse_response = [0.1 + i for i in range(8)]
+ impulse_response += reversed(impulse_response)
+ sfg = symmetric_fir(impulse_response)
+ assert (
+ len(
+ [
+ comp
+ for comp in sfg.components
+ if isinstance(comp, ConstantMultiplication)
+ ]
+ )
+ == 8
+ )
+ assert len([comp for comp in sfg.components if isinstance(comp, Addition)]) == 15
+ assert len([comp for comp in sfg.components if isinstance(comp, Delay)]) == 15
+
+ sim = Simulation(sfg, [Impulse()])
+ sim.run_for(17)
+ impulse_response.append(0.0)
+ assert np.allclose(sim.results["0"], impulse_response)
+
+ impulse_response = [0.1 + i for i in range(50)]
+ impulse_response += reversed(impulse_response)
+ sfg = symmetric_fir(impulse_response)
+ assert (
+ len(
+ [
+ comp
+ for comp in sfg.components
+ if isinstance(comp, ConstantMultiplication)
+ ]
+ )
+ == 50
+ )
+ assert len([comp for comp in sfg.components if isinstance(comp, Addition)]) == 99
+ assert len([comp for comp in sfg.components if isinstance(comp, Delay)]) == 99
+
+ sim = Simulation(sfg, [Impulse()])
+ sim.run_for(101)
+ impulse_response.append(0.0)
+ assert np.allclose(sim.results["0"], impulse_response)
+
+ with pytest.raises(ValueError, match="Coefficients must be of even length"):
+ symmetric_fir([0.1, 0.2, 0.1])
+
+
def test_sfg_generator_errors():
- sfg_gens = [wdf_allpass, transposed_direct_form_fir, direct_form_fir]
+ sfg_gens = [wdf_allpass, transposed_direct_form_fir, direct_form_fir, symmetric_fir]
for gen in sfg_gens:
with pytest.raises(ValueError, match="Coefficients cannot be empty"):
gen([])