def test_blandAtlman_axis_handle(self): import matplotlib with self.subTest(msg='Single axis'): fig, ax = matplotlib.pyplot.subplots() ax = pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', savePath=None, ax=ax) self.assertTrue(isinstance(ax, matplotlib.axes._subplots.Axes)) with self.subTest(msg='Multiple axes'): fig, ax = matplotlib.pyplot.subplots(2,2) ax1 = pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', savePath=None, ax=ax[1,1]) ax2 = pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', savePath=None, ax=ax[0,0]) self.assertTrue(isinstance(ax1, matplotlib.axes._subplots.Axes)) self.assertTrue(isinstance(ax2, matplotlib.axes._subplots.Axes))
def test_blandAtlman_screen(self):
noSamp = numpy.random.randint(100, high=500, size=None)
pyCompare.blandAltman(numpy.random.rand(noSamp) * 100 + 100,
numpy.random.rand(noSamp) * 50 + 100,
confidenceIntervalMethod='approximate',
savePath=None)
def test_blandAtlman_saves(self): with tempfile.TemporaryDirectory() as tmpdirname: with self.subTest(msg='Default Parameters'): outputPath = os.path.join(tmpdirname, 'plot') pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', savePath=outputPath) self.assertTrue(os.path.exists(outputPath)) with self.subTest(msg='No CIs'): outputPath = os.path.join(tmpdirname, 'noCIplot') pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceInterval=None, savePath=outputPath) self.assertTrue(os.path.exists(outputPath)) with self.subTest(msg='Percentage'): outputPath = os.path.join(tmpdirname, 'plot_percentage') pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', percentage=True, savePath=outputPath) self.assertTrue(os.path.exists(outputPath))
def test_blandAtlman_screen(self): pyCompare.blandAltman(numpy.random.rand(self.noSamp)*100+100, numpy.random.rand(self.noSamp)*50+100, confidenceIntervalMethod='approximate', savePath=None)
tempt_pulsatile = pulsatile.iloc[test_list.iloc[0].values.squeeze().tolist()]
tempt_pulsatile = (tempt_pulsatile-tempt_pulsatile.mean())/tempt_pulsatile.std()
tempt_pulsatile.iloc[0].to_numpy()
test_predict[0][0]
test_pred_df = []
test_true_df = []
for item in range(10):
for idx in range(19):
test_pred_df.extend(test_predict[item][idx])
tempt_pulsatile = pulsatile.iloc[test_list.iloc[item].values.squeeze().tolist()]
tempt_pulsatile = (tempt_pulsatile-tempt_pulsatile.mean())/tempt_pulsatile.std()
test_true_df.extend(tempt_pulsatile.iloc[idx].to_numpy())
blandAltman(test_pred_df, test_true_df,savePath=os.path.join(run_name, 'BA_model1.png'),figureFormat='png')
test_predict[0]
def dice(im1, im2):
"""
Computes the Dice coefficient, a measure of set similarity.
Parameters
----------
im1 : array-like, bool
Any array of arbitrary size. If not boolean, will be converted.
im2 : array-like, bool
Any other array of identical size. If not boolean, will be converted.
Returns
-------
dice : float
# verbose=0)
out = model.predict(X_train2, batch_size=batch_size)
predicted2 = out.ravel()
out = model.predict(X_test, batch_size=batch_size)
predicted = out.ravel()
# sns.set(color_codes=True)
# ax = sns.regplot(x=Y_test2, y=predicted2, color="g")
r1 = r2_score(Y_train2, predicted2)
r2 = r2_score(Y_test, predicted)
print("New Signal R2:{0}".format(r1))
print("Same Signal R2:{0}".format(r2))
plt.plot(Y_train2)
plt.plot(predicted2)
plt.legend(['Target', 'Estimated'], loc='upper left')
plt.show()
blandAltman(Y_test,
predicted,
savePath='SavedFigureAltman.svg',
figureFormat='svg')
df = pd.DataFrame(columns=["target", "predicted"])
df["predicted"] = predicted2
df["target"] = Y_train2
df.to_excel("result.xlsx")
pass
import urllib.request, json
import pyCompare
url = "https://cvdhd-ef484.firebaseio.com/.json"
data = urllib.request.urlopen(url).read().decode()
obj = json.loads(data)
print(obj)
bmp = []
dht = []
for key in obj:
bmp.append(obj[key]["bmp"]["temp"])
dht.append(obj[key]["dht"]["temp"])
print(bmp)
print(dht)
pyCompare.blandAltman(bmp,dht)
"""
Myocardial Mass ED
"""
auto_ED_Myo_Mass = np.array([
183349.609375, 181792.96875, 182976.822669983, 60789.55078125,
156115.72265625, 157034.118652344, 95858.3605499268, 125610.3515625,
146155.725517273, 66088.8671875, 105493.1640625, 95166.015625,
102083.472846985, 94833.6481323242, 73320.6992340088
])
manual_ED_Myo_Mass = np.array([
164257.8125, 183691.40625, 188135.83820343, 70459.716796875,
176676.940917969, 187925.354003906, 92289.4210891724, 137386.322021484,
132912.886299133, 65502.9296875, 98706.0546875, 106298.828125,
97772.5845108032, 93511.6177215576, 68698.119140625
])
pyCompare.blandAltman(auto_ED_Myo_Mass,
manual_ED_Myo_Mass,
limitOfAgreement=1.96,
confidenceInterval=95,
title='Myo Mass ED (g) rho=0.96',
meanColour='c',
loaColour='gray',
pointColour='red',
dpi=1000,
savePath='bland_MyoMassED.png')
