From 12c5a23911d2a2b01b3103d5ac5b9db0922bcb2a Mon Sep 17 00:00:00 2001
From: Andreas Kloeckner <inform@tiker.net>
Date: Wed, 5 Jun 2013 01:50:23 -0400
Subject: [PATCH] PEP8 test_linalg.
---
test/test_linalg.py | 120 +++++++-------------------------------------
1 file changed, 18 insertions(+), 102 deletions(-)
diff --git a/test/test_linalg.py b/test/test_linalg.py
index 0bd22021a..15073433d 100644
--- a/test/test_linalg.py
+++ b/test/test_linalg.py
@@ -23,40 +23,16 @@ THE SOFTWARE.
"""
-
-
import numpy as np
-import numpy.linalg as la
import pyopencl as cl
import pyopencl.array as cl_array
import loopy as lp
-from pyopencl.tools import pytest_generate_tests_for_pyopencl \
- as pytest_generate_tests
-
-
-
-
-def make_well_conditioned_dev_matrix(queue, shape, dtype=np.float32,
- order="C", ran_factor=1, id_factor=5, inc_factor=0, od=0):
- if isinstance(shape, int):
- shape = (shape, shape)
- l = max(shape)
- eye_ish = id_factor*np.eye(l, k=od)
- if inc_factor:
- eye_ish[np.arange(l), np.arange(l)] = inc_factor*np.arange(l)
- ary = np.asarray(
- ran_factor*np.random.randn(*shape)
- + eye_ish[:shape[0], :shape[1]],
- dtype=dtype, order=order)
-
- return cl_array.to_device(queue, ary)
+from pyopencl.tools import ( # noqa
+ pytest_generate_tests_for_pyopencl
+ as pytest_generate_tests)
-
-
-DO_CHECK = True
-
DEBUG_PREAMBLE = r"""
#pragma OPENCL EXTENSION cl_amd_printf: enable
#define MY_J (j_outer*64+j_inner_outer*16+j_inner_inner)
@@ -64,43 +40,10 @@ DEBUG_PREAMBLE = r"""
#define IFDIAG if (MY_I == MY_J)
#define TST(S) if (MY_J == 144 && MY_I == 16-48) \
for (int aa = 0; aa < 16: ++ab) \
- for (int bb = 0; bb < 16: ++bb)
+ for (int bb = 0; bb < 16: ++bb)
"""
-
-
-def check_error(refsol, sol):
- if not DO_CHECK:
- return
-
- if sol.shape == 2:
- norm_order = "fro"
- else:
- norm_order = 2
-
- rel_err = la.norm(refsol-sol, norm_order)/la.norm(refsol, norm_order)
- if rel_err > 1e-5 or np.isinf(rel_err) or np.isnan(rel_err):
- if 1:
- import matplotlib.pyplot as pt
- pt.imshow(refsol-sol, interpolation="nearest")
- pt.colorbar()
- pt.show()
- elif 0:
- print "---------------------------"
- print "ACTUAL"
- print "---------------------------"
- np.set_printoptions(threshold=1000000, linewidth=200)
- print sol[:16,:16]
- print "---------------------------"
- print "CORRECT"
- print "---------------------------"
- print refsol[:16,:16]
- raise RuntimeError("check failed, rel err=%g" % rel_err)
-
-
-
-
def get_suitable_size(ctx):
dev, = ctx.devices
if dev.type == cl.device_type.CPU:
@@ -109,11 +52,11 @@ def get_suitable_size(ctx):
return 1600
-
-
def check_float4(result, ref_result):
for comp in ["x", "y", "z", "w"]:
- return np.allclose(ref_result[comp], result[comp], rtol=1e-3, atol=1e-3), None
+ return np.allclose(
+ ref_result[comp], result[comp], rtol=1e-3, atol=1e-3), None
+
def test_axpy(ctx_factory):
ctx = ctx_factory()
@@ -155,8 +98,10 @@ def test_axpy(ctx_factory):
def variant_gpu(knl):
unroll = 4
block_size = 256
- knl = lp.split_iname(knl, "i", unroll*block_size, outer_tag="g.0", slabs=(0, 1))
- knl = lp.split_iname(knl, "i_inner", block_size, outer_tag="unr", inner_tag="l.0")
+ knl = lp.split_iname(knl, "i", unroll*block_size,
+ outer_tag="g.0", slabs=(0, 1))
+ knl = lp.split_iname(knl, "i_inner", block_size,
+ outer_tag="unr", inner_tag="l.0")
return knl
for variant in [variant_cpu, variant_gpu]:
@@ -170,9 +115,6 @@ def test_axpy(ctx_factory):
parameters={"a": a, "b": b, "n": n}, check_result=check)
-
-
-
def test_transpose(ctx_factory):
dtype = np.dtype(np.float32)
ctx = ctx_factory()
@@ -207,8 +149,6 @@ def test_transpose(ctx_factory):
parameters={})
-
-
def test_plain_matrix_mul(ctx_factory):
ctx = ctx_factory()
order = "C"
@@ -249,9 +189,6 @@ def test_plain_matrix_mul(ctx_factory):
parameters={"n": n}, check_result=check)
-
-
-
def test_variable_size_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -291,9 +228,6 @@ def test_variable_size_matrix_mul(ctx_factory):
parameters={"n": n})
-
-
-
def test_rank_one(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -374,8 +308,6 @@ def test_rank_one(ctx_factory):
parameters={"n": n})
-
-
def test_troublesome_premagma_fermi_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -416,8 +348,6 @@ def test_troublesome_premagma_fermi_matrix_mul(ctx_factory):
parameters={})
-
-
def test_intel_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -455,7 +385,8 @@ def test_intel_matrix_mul(ctx_factory):
# FIXME: Grouped prefetch
#knl = lp.add_prefetch(knl, 'a', ["k_inner", ("i_inner_inner", "i_inner_outer")])
- #knl = lp.add_prefetch(knl, 'b', ["k_inner", ("j_inner_inner", "j_inner_outer"),])
+ #knl = lp.add_prefetch(knl, 'b',
+ # ["k_inner", ("j_inner_inner", "j_inner_outer"),])
kernel_gen = lp.generate_loop_schedules(knl)
#hints=["k_outer", "k_inner_outer", "k_inner_inner"]
@@ -466,9 +397,6 @@ def test_intel_matrix_mul(ctx_factory):
parameters={})
-
-
-
def test_magma_fermi_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -495,7 +423,6 @@ def test_magma_fermi_matrix_mul(ctx_factory):
i_chunks = 16
j_chunks = 16
-
knl = lp.split_iname(knl, "i", i_reg*i_chunks, outer_tag="g.0")
knl = lp.split_iname(knl, "i_inner", i_reg, outer_tag="l.0", inner_tag="ilp")
knl = lp.split_iname(knl, "j", j_reg*j_chunks, outer_tag="g.1")
@@ -504,7 +431,8 @@ def test_magma_fermi_matrix_mul(ctx_factory):
knl = lp.split_iname(knl, "k_inner", 8, outer_tag="unr")
# FIXME
#knl = lp.add_prefetch(knl, 'a', ["k_inner", "i_inner_inner", "i_inner_outer"])
- #knl = lp.add_prefetch(knl, 'b', ["k_inner", ("j_inner_inner", "j_inner_outer"),])
+ #knl = lp.add_prefetch(knl, 'b',
+ # ["k_inner", ("j_inner_inner", "j_inner_outer"),])
kernel_gen = lp.generate_loop_schedules(knl)
kernel_gen = lp.check_kernels(kernel_gen, dict(n=n))
@@ -514,9 +442,6 @@ def test_magma_fermi_matrix_mul(ctx_factory):
parameters={})
-
-
-
def test_image_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -553,7 +478,6 @@ def test_image_matrix_mul(ctx_factory):
parameters={})
-
def test_image_matrix_mul_ilp(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -579,7 +503,8 @@ def test_image_matrix_mul_ilp(ctx_factory):
knl = lp.split_iname(knl, "i", 2, outer_tag="g.0", inner_tag="l.1")
j_inner_split = 4
knl = lp.split_iname(knl, "j", ilp*j_inner_split, outer_tag="g.1")
- knl = lp.split_iname(knl, "j_inner", j_inner_split, outer_tag="ilp", inner_tag="l.0")
+ knl = lp.split_iname(knl, "j_inner", j_inner_split,
+ outer_tag="ilp", inner_tag="l.0")
knl = lp.split_iname(knl, "k", 2)
# conflict-free?
knl = lp.add_prefetch(knl, 'a', ["i_inner", "k_inner"])
@@ -593,7 +518,6 @@ def test_image_matrix_mul_ilp(ctx_factory):
parameters={})
-
def test_ilp_race_matmul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -623,9 +547,6 @@ def test_ilp_race_matmul(ctx_factory):
list(lp.generate_loop_schedules(knl))
-
-
-
def test_fancy_matrix_mul(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -650,7 +571,7 @@ def test_fancy_matrix_mul(ctx_factory):
knl = lp.split_iname(knl, "i", 16, outer_tag="g.0", inner_tag="l.1")
knl = lp.split_iname(knl, "j", 16, outer_tag="g.1", inner_tag="l.0")
- knl = lp.split_iname(knl, "k", 16, slabs=(0,1))
+ knl = lp.split_iname(knl, "k", 16, slabs=(0, 1))
knl = lp.add_prefetch(knl, 'a', ["i_inner", "k_inner"])
knl = lp.add_prefetch(knl, 'b', ["k_inner", "j_inner"])
@@ -662,9 +583,6 @@ def test_fancy_matrix_mul(ctx_factory):
parameters=dict(n=n))
-
-
-
def test_small_batched_matvec(ctx_factory):
dtype = np.float32
ctx = ctx_factory()
@@ -702,8 +620,6 @@ def test_small_batched_matvec(ctx_factory):
parameters=dict(K=K))
-
-
if __name__ == "__main__":
import sys
if len(sys.argv) > 1:
--
GitLab