def test_Exp1D_double(self):
npr = np.exp(self.np_double_a)
dcr = dc.exp(self.dc_double_a)
np.testing.assert_allclose(npr,
np.array(dcr.data()).astype(np.float64),
rtol=1e-3,
atol=1e-3)
def test_Exp2D_double_2(self):
np_double_a = np.reshape(self.np_double_a, (6, 8))
dc_double_a = dc.reshape(self.dc_double_a, (6, 8))
npr = np.exp(np_double_a)
dcr = dc.exp(dc_double_a)
np.testing.assert_allclose(npr.flatten(),
np.array(dcr.data()).astype(np.float64),
rtol=1e-3,
atol=1e-3)
def test_Exp2D_float_3(self):
np_float_a = np.reshape(self.np_float_a, (12, 4))
dc_float_a = dc.reshape(self.dc_float_a, (12, 4))
npr = np.exp(np_float_a)
dcr = dc.exp(dc_float_a)
np.testing.assert_allclose(npr.flatten(),
np.array(dcr.data()).astype(np.float32),
rtol=1e-3,
atol=1e-3)
result = run(command,
stdout=PIPE,
stderr=PIPE,
universal_newlines=True,
shell=True)
for line in result.stdout.split("\n"):
if (line.find("writing file ") == 0):
resultFile = line[13:line.find('.', -1)].split()[0]
with open(resultFile, 'r') as f:
return f.read()
return ""
# Run model in the loop
import deepC.dnnc as dc
for i in range(5):
index = random.randint(0, len(images) - 1)
write_image(index)
model_result = run_model("./mnist.exe ./image.data").strip("[]")
# Convert log softmax output to probability
log_probs = dc.array([float(f) for f in model_result.strip("[]").split()])
probabilities = dc.exp(log_probs)
trueLabel = labels[index]
prediction = dc.argmax(probabilities)[0]
display(images[index])
print("True label = ", labels[index])
print("Model Prediction: ", dc.argmax(probabilities))
