# 导入数据集获取工具# from sklearn.datasets import fetch_mldata# 加载MNIST手写数字数据集# mnist=fetch_mldata('MNIST original')# 报错,无法获取,参考: /Dontla/article/details/99677247
# 获取scikit数据根目录from sklearn.datasets import get_data_homeprint(get_data_home())
C:\Users\HuaWei\scikit_learn_data
from sklearn.datasets import fetch_mldatamnist=fetch_mldata('MNIST original')mnist
c:\users\huawei\appdata\local\programs\python\python36\lib\site-packages\sklearn\utils\deprecation.py:85: DeprecationWarning: Function fetch_mldata is deprecated; fetch_mldata was deprecated in version 0.20 and will be removed in version 0.22. Please use fetch_openml.warnings.warn(msg, category=DeprecationWarning)c:\users\huawei\appdata\local\programs\python\python36\lib\site-packages\sklearn\utils\deprecation.py:85: DeprecationWarning: Function mldata_filename is deprecated; mldata_filename was deprecated in version 0.20 and will be removed in version 0.22. Please use fetch_openml.warnings.warn(msg, category=DeprecationWarning){'DESCR': ' dataset: mnist-original','COL_NAMES': ['label', 'data'],'target': array([0., 0., 0., ..., 9., 9., 9.]),'data': array([[0, 0, 0, ..., 0, 0, 0],[0, 0, 0, ..., 0, 0, 0],[0, 0, 0, ..., 0, 0, 0],...,[0, 0, 0, ..., 0, 0, 0],[0, 0, 0, ..., 0, 0, 0],[0, 0, 0, ..., 0, 0, 0]], dtype=uint8)}
import numpy as npimport matplotlib.pyplot as pltprint('type(mnist):',type(mnist),'\n')print('样本数量:{}\n样本特征数:{}'.format(mnist.data.shape[0],mnist.data.shape[1]),'\n')print('mnist.data[0].shape:',mnist.data[0].shape)
type(mnist): <class 'sklearn.utils.Bunch'> 样本数量:70000样本特征数:784 mnist.data[0].shape: (784,)
# 打印数字来看看x=np.arange(-14,14,1)xx,yy=np.meshgrid(x,x)# 千位决定数字几z=mnist.data[62957].reshape(xx.shape)plt.figure()# 先将z转置,再逆时针旋转90度plt.pcolormesh(xx,yy,np.rot90(z.T,1))
<matplotlib.collections.QuadMesh at 0x2639a419f98>
# 建立训练数据集和测试数据集from sklearn.model_selection import train_test_splitX=mnist.data/255print('打印mnist.data:\n',mnist.data,'\n')print('mnist.data的维度:\n',(mnist.data).shape)y=mnist.targetX_train,X_test,y_train,y_test=train_test_split(X,y,train_size=5000,test_size=1000,random_state=62)
打印mnist.data:[[0 0 0 ... 0 0 0][0 0 0 ... 0 0 0][0 0 0 ... 0 0 0]...[0 0 0 ... 0 0 0][0 0 0 ... 0 0 0][0 0 0 ... 0 0 0]] mnist.data的维度:(70000, 784)
# 训练MLP神经网络# 设置神经网络有两个100节点的隐藏层from sklearn.neural_network import MLPClassifiermlp_hw=MLPClassifier(solver='lbfgs',hidden_layer_sizes=[100,100],activation='relu',alpha=1e-5,random_state=62)# 使用数据集寻来你神经网络模型mlp_hw.fit(X_train,y_train)print('测试数据得分:{:.2f}%'.format(mlp_hw.score(X_test,y_test)*100))
测试数据得分:93.60%
# 先安装PIL包, pip install pillowfrom PIL import Imageimage=Image.open('3.jpg').convert('F')# 调整图像的大小image=image.resize((28,28))arr=[]# 将图像中的像素作为预测数据点的特征for i in range(28):for j in range(28):pixel=1.0-float(image.getpixel((j,i)))/255arr.append(pixel)# 由于只有一个样本,所以需要进行reshape操作arr1=np.array(arr).reshape(1,-1) # reshape成一行, 无论多少列# 进行图像识别print("图片中的数字是:{:.0f}".format(mlp_hw.predict(arr1)[0]))
图片中的数字是:3
# 将图片获取像素再把它画出来看看image=Image.open('汉.jpg').convert('F')# 调整图像的大小image=image.resize((28,28))x=np.arange(-14,14,1)xx,yy=np.meshgrid(x,x)arr=[]for i in range(28):for j in range(28):pixel=image.getpixel((j,27-i))arr.append(pixel)arr_=np.array(arr)print(arr_)z=arr_.reshape(xx.shape)plt.figure()# 顺时针旋转90度plt.pcolormesh(xx,yy,z)
[207.7539978 210.98199463 213.03599548 216.92599487 220.76699829222.66299438 225.64599609 229.31900024 228.94900513 232.70399475234.68699646 236.1210022 236.40899658 237.81100464 237.76199341234.76199341 234.66499329 232.8500061 228.8500061 229.06300354225.07400513 228.1190033 221.34300232 220.022995 213.92599487215.11099243 214.96499634 209.0039978 209.7539978 210.98199463212.96499634 215.86099243 219.76699829 220.66299438 226.1190033226.09100342 231.70399475 231.70399475 210.75300598 236.60499573238.1210022 238.1210022 238.32699585 235.03900146 235.13800049232.8500061 230.8500061 229.06300354 226.74299622 224.16200256222.34300232 220.022995 216.86599731 158.18499756 210.73699951210.0039978 210.76800537 210.91099548 212.11099243 193.69599915220.65299988 221.66299438 224.1190033 228.10 229.70399475232.70399475 234.8500061 224.16000366 236.8500061 236.1210022235.32699585 230.32699585 218.23699951 237.70399475 228.85227.09100342 221.13000488 204.69599915 206.8769989 170.13499451216.63800049 153.21600342 208.79299927 207.14599609 165.83599854208.39100647 163.04299927 161.19900513 219.79499817 222.95100403223.64599609 205.22900391 231.70399475 232.70399475 232.8500061234.96400452 205.9750061 215.76199341 236.32699585 234.09899902216.8500061 236.26899719 233.30200195 199.64599609 227.08000183232.897995 222.78399658 190.8769989 216.81199646 154.93800354199.39100647 210.06399536 209.91099548 151.875212.93299866172.8500061 221.93299866 222.90800476 224.0217 175.22900391228.75300598 230.98100281 233.8500061 236.04400635 51.85400009251.08200073 238.08200073 236.03900146 234.8500061 227.06300354227.76400757 225.74299622 225.94299316 70.76399994 226.36099243211.2749939 218.15400696 217.23199463 205.88299561 198.70799255205.09100342 202.28900146 212.96499634 164.07299805 219.91900635221.1190033 217.86399841 179.22900391 142.95300293 233.13800049236.8500061 233.66499329 230.03900146 46.77199936 37.08200073236.13800049 234.95300293 233.75300598 232.16000366 231.7039947543.70700073 47.07099915 230.27999878 219.88299561 218.15400696219.11300659 216.77600098 214.30900574 206.7539978 151.76800537215.01400757 163.66900635 216.65299988 214.76199341 226.34300232160.88600159 50.91799927 46. 232.8500061 241.96400452218.1210022 246.85400391 49.2879982 200.03900146 220.70399475231.13800049 208.85 52.81100082 47.08200073 208.80299377218.73500061 220.05499268 218.15400696 212.92599487 211.03599548212.37399292 203.98199463 216.38699341 210.80099487 165.82299805220.34300232 220.66299438 225.64599609 198.95100403 47.47. 237.40899658 233.60499573 234.18099976 238.7940063556.08200073 47.16400146 208.94900513 213.72099304 142.9810028145.68999863 50.04399872 180.14399719 222.11599731 218.05499268219.92599487 214.11099243 214.01499939 208.94599915 207.99899292210.7539978 213.727005 186.71800232 220.65299988 223.27799988226.88499451 224.82000732 56.25 46.04899979 46.94200134238.70399475 215.94200134 235.08200073 237.8710022 47.31000137202.80299377 46.7120018 174.18099976 231.85 226.07400513224.94099426 220.86599731 217.8223 216.86599731 215.83399963208.01499939 211.24299622 210.7539978 212.69400024 162.14599609216.80099487 220.99499512 223.13000488 232.22900391 202.216261.22999954 225.95300293 46.31000137 46.31000137 49.3100013746.77199936 58.08200073 47.31000137 182.62199402 231.00900269231.96000671 228.80299377 225.08500671 222.2440033 220.91499329219.10499573 217.86599731 213.83399963 211.96499634 152.9960022210.05299377 212.91099548 198.17999268 216.8769989 219.99499512224.90800476 208.05000305 226.80299377 230.06300354 233.9810028146.85400009 234.82800293 217.03900146 243.03900146 234.08200073219.15299988 46.16799927 43.79399872 48.91799927 49.1459999147.1570015 46.7120018 224.74499512 220.01199341 219.15400696215.83399963 212.63800049 210.82499695 209.98199463 212.98199463159.03900146 215.23599243 219.76699829 223.65299988 223.88499451228.03100586 229.06300354 232.70399475 204.91000366 208.91000366240.13800049 233.32699585 52.32699966 51.85400009 47.08200073224.13800049 211.72099304 229.06300354 211.31599426 179.55599976211.8769989 202.96499634 216.86599731 216.83399963 209.76600647209.0039978 208.98199463 210.98199463 160.03599548 218.30000305220.76699829 224.9019928 230.1190033 189.21800232 243.72099304231.8500061 200.77000427 215.16499329 237.18099976 238.09899902224.05000305 236.91000366 47.45.77199936 46.95100021220.16000366 231.08500671 164.74499512 187.8769989 219.17599487217.86599731 213.83399963 163.72900391 202.30900574 207.98199463205.06399536 143.1519928 219.0723 216.76699829 221.66299438208.00100708 190.0723 232.11300659 230.8500061 215.8500061230.98100281 235.18099976 83.08200073 49.32699966 47.0880012544.77199936 235.13800049 232.04100037 224.97099304 225.74299622222.66299438 225.94099426 226.19299316 217.15400696 213.83399963205.91499329 210.92199707 209.02799988 208.69400024 210.82499695215.09399414 219.76699829 220.66299438 224.44000244 225.82000732230.19799805 232.8500061 234.96400452 202.8500061 208.1809997647.08200073 68. 182.97099304 27.85400009 45.8320007345.89099884 69.11599731 225.82499695 222.9019928 224.21800232219.90800476 212.65299988 170.76800537 205.90800476 207.2162206.83599854 169.66900635 215.92199707 216.91499329 220.65299988223.0403 46.7120018 47.47.1570015 33.72900009234.66499329 222.8500061 239.18099976 39.07099915 49.287998278.82800293 41.7120018 230.86099243 29.93199921 46.8689994847.02199936 226.19500732 222.0059967 219.77799988 217.86599731216.11099243 211.69400024 169.70799255 204.7539978 203.0059967210.13499451 215.06399536 217.30000305 197.17100525 45.07099915230.91000366 223.62199402 232.98100281 234.81100464 232.8500061236.147995 103.32700348 238.32699585 234.87600708 42.8540000926.72900009 44.02199936 43.87099838 226.08000183 211.09399414223.09899902 220.022995 216.86599731 216.11099243 218.01499939201.97399902 211.7539978 211.96499634 144.01300049 192.89399719219.76699829 223.03300476 230.92399597 226.05499268 214.2440033237.06300354 229.75300598 234.8500061 237.40899658 238.32699585199.8710022 47.31000137 47.31000137 75.03900146 232.30200195230.06300354 227.10800171 199.78399658 217.2945 214.79499817216.86599731 213.86599731 212.3769989 210.05299377 210.05299377208.13499451 218.66000366 214.80099487 219.99499512 224.91900635222.90800476 224.1190033 211.88000488 234.95300293 235.32699585239.03900146 233.16999817 238.08200073 93.49.31000137222.03900146 47.08200073 64.76200104 227.30200195 228.10800171195.94099426 182.03700256 220.73500061 216.63800049 213.81199646206.82600403 190.66900635 207.7539978 193.0039978 154.28399658219.9384 219.99499512 224.0217 228.13000488 212.1190033229.94299316 248.32699585 47.31000137 47.31000137 46.0820007345.07099915 47.49.31000137 45.72900009 43.0820007347.32699966 225.9980011 227.88000488 223.26100159 183.79299927219.022995 214.92599487 210.12199402 210.02099609 180.06399536208.7539978 211.71099854 166.87199402 211.17599487 219.99499512220.95100403 224.64599609 226.78100586 86.16400146 48.9830017147.31000137 40.03900146 239.40899658 233.08200073 47.50. 228.05000305 43.08200073 47.31000137 45.3269996651.66500092 226.44000244 185.25300598 217.11099243 214.15400696214.11099243 220.65499878 153.83999634 207.7539978 151.94599915214.10299683 215.19299316 219.77799988 192.13499451 219.9019928166.91499329 230.9384 225.1210022 232.18099976 237.95300293236.40899658 240.40899658 47. 231.15299988 232.22000122206.1289978 47.31000137 127.13800049 226.93400574 223.0403222.9019928 219.79499817 218.15400696 216.11099243 178.8500061149.95500183 201.22599792 209.98199463 214.25300598 217.0723220.05499268 226.90800476 223.9019928 227.13000488 192.23899841233.70399475 232.9384 216.8500061 234.1210022 236.8500061240.72900391 234.76199341 234.8500061 239.91000366 47.3100013749.95100021 224.03100586 205.25300598 216.15400696 218.11099243216.86599731 215.67700195 209.9939 212.95199585 204.29400635212.02099609 214.12199402 215.36000061 220.01199341 223.13000488226.10 189.55599976 233.88000488 232.70399475 235.86099243214.70399475 216.95300293 238.03900146 238.05000305 208.66499329234.9750061 235.96400452 43.85400009 248.17199707 223.04499817214.94099426 225.92399597 217.11099243 217.86599731 151.83900452140.05099487 210.71099854 208.11300659 157.40499878 156.02799988204.92599487 220.99499512 221.26100159 189.21800232 198.08900452226.96000671 232.70399475 201.80299377 216.16000366 234.60499573235.81100464 237.72900391 201.26899719 242.23699951 224.93400574226.80799866 229.06300354 218.08799744 222.30000305 212.27600098213.88299561 215.92599487 215.00100708 144.06700134 211.33099365208.08599854 157.91099548 212.04699707 226.03599548 219.99499512223.0217 224.82000732 201.30000305 200.23699951 233.06300354235.13800049 212.75300598 238.16999817 235.81100464 236.08799744238.98100281 235.16000366 200.2440033 230.70399475 229.06300354225.05499268 174.95500183 222.34300232 201.67700195 216.63800049215.19299316 160.19900513 165.22399902 208.7539978 149.33700562213.11099243 168.67700195 219.76699829 222.82000732 226.89100647232.1190033 230.23699951 231.8500061 194.75300598 204.70399475234.89300537 238.03900146 238.1210022 234.60499573 234.86099243204.1190033 233.94299316 230.06300354 227.2162 168.16799927214.95599365 217.01199341 216.63800049 215.83399963 170.66900635196.28399658 209.03100586 211.67199707 215.63800049 217.09399414219.15400696 219.98399353 224.1190033 226.96000671 228.72099304232.9939 232.96400452 237.18099976 233.1210022 238.03900146238.03900146 237.09899902 235.1210022 229.9384 231.06300354230.06300354 226.85 229.84199524 226.34300232 187.98699951217.79499817 214.83399963 212.92199707 204.7539978 ]<matplotlib.collections.QuadMesh at 0x263b6ca90b8>
# 检测reshape()函数机制x=np.array([1,2,3,4,5,6,7,8,9])y=np.array([[1,1,2],[1, 2, 1],[1, 1, 1]])x=x.reshape(y.shape)print(x)
[[1 2 3][4 5 6][7 8 9]]
# 将四色图片获取像素再把它画出来看看image=Image.open('四色.jpg').convert('F')# image=image.resize((16,16))x=np.arange(-10,10,1)xx,yy=np.meshgrid(x,x)arr=[]# PIL Image似乎是以图像左下角为原点的for i in range(20):for j in range(20):pixel=image.getpixel((j,19-i))arr.append(pixel)arr_=np.array(arr)# print(arr_)z=arr_.reshape(xx.shape)print(z)plt.figure()# 顺时针旋转90度plt.pcolormesh(xx,yy,z)
[[2.52750000e+02 2.53205994e+02 2.53516006e+02 2.52330994e+022.5496e+02 2.51919998e+02 2.54070999e+02 2.54412994e+022.51951004e+02 2.53826004e+02 1.12500000e+00 8.97000015e-013.84299994e+00 2.98999995e-01 2.98999995e-01 5.27000010e-010.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.53048996e+02 2.52917999e+02 2.52341995e+02 2.51000000e+022.51921997e+02 2.51009995e+02 2.51807999e+02 2.52149994e+022.52035995e+02 2.51694000e+02 6.81099987e+00 5.92500019e+003.55500007e+00 3.55500007e+00 3.55500007e+00 5.02799988e+008.85999978e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.54412994e+02 2.54412994e+02 2.53238998e+02 2.52947998e+022.47134003e+02 2.44854004e+02 2.45651993e+02 2.45994003e+022.45880005e+02 2.45423996e+02 9.76799965e+00 9.18099976e+008.88199997e+00 1.06540003e+01 1.03549995e+01 7.09899998e+002.36999989e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.54412994e+02 2.53826004e+02 2.53826004e+02 2.52834000e+022.16563995e+02 2.17845001e+02 2.18656006e+02 2.17738007e+022.15895004e+02 2.17895004e+02 2.19026001e+02 2.17009003e+022.18718994e+02 2.18007004e+02 2.16867004e+02 2.19136002e+022.95700002e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.53804001e+02 2.54701004e+02 2.5377e+02 2.53860001e+022.52606003e+02 2.49755997e+02 2.45651993e+02 2.16949005e+022.45651993e+02 2.46677994e+02 9.45800018e+00 7.09899998e+005.62599993e+00 4.44099998e+00 2.18643997e+02 5.61499977e+002.07100010e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.54701004e+02 2.51287994e+02 2.54544006e+02 2.53843002e+022.54412994e+02 2.52149994e+02 2.45994003e+02 2.0001e+022.48388000e+02 2.12800995e+02 2.13563995e+02 8.85999966e+002.36999989e+00 1.48399997e+00 4.44099998e+00 4.72900009e+008.85999978e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.54412994e+02 2.51082001e+02 2.53255997e+02 2.53957001e+022.54070999e+02 2.48363007e+02 2.45651993e+02 2.13779007e+022.50781998e+02 2.51123993e+02 2.7997e+02 8.84899998e+005.01700020e+00 2.98999995e-01 1.19599998e+00 1.19599998e+002.98999995e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.52897003e+02 2.53598007e+02 2.53957001e+02 2.50854004e+022.5477e+02 2.5135e+02 2.43942001e+02 2.18639008e+022.51123993e+02 2.51807999e+02 2.15600006e+02 1.09309998e+016.0014e+00 4.72900009e+00 1.78299999e+00 5.02799988e+002.98999995e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.5277e+02 2.5477e+02 2.51059998e+02 2.54059998e+022.53000000e+02 2.50781998e+02 2.4998e+02 2.20854004e+022.50554001e+02 2.50326004e+02 2.17130997e+02 2.21694000e+022.17798004e+02 2.20908005e+02 2.15089005e+02 4.13100004e+008.85999978e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.54473007e+02 2.54059998e+02 2.54701004e+02 2.54287994e+022.54701004e+02 2.51807999e+02 2.15817993e+02 2.17177002e+022.48792999e+02 2.50326004e+02 2.15914001e+02 1.53940001e+011.44969997e+01 8.58300018e+00 5.90299988e+00 3.23399997e+001.17400002e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.69000006e+00 5.87000012e-01 5.27000010e-01 7.54999995e-011.04299998e+00 3.53299999e+00 3.55500007e+00 4.45200014e+005.32700014e+00 3.85400009e+00 2.19878006e+02 1.21379995e+016.22399998e+00 4.44099998e+00 2.05999994e+00 3.83200002e+005.87000012e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][5.87000012e-01 1.28799999e+00 2.17899990e+00 1.46700001e+001.04299998e+00 1.40199995e+00 2.00000000e+00 2.98999995e-011.48399997e+00 6.21299982e+00 2.16121002e+02 9.46899986e+005.03900003e+00 2.36999989e+00 2.35899997e+00 5.87000012e-010.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][5.87000012e-01 2.00000000e+00 1.76600003e+00 1.35300004e+001.22800004e+00 5.87000012e-01 1.17400002e+00 1.17400002e+002.95700002e+00 5.32700014e+00 2.16294998e+02 1.24150000e+016.51200008e+00 3.54399991e+00 1.18499994e+00 7.97999978e-012.28000000e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.07100010e+00 2.98999995e-01 2.28000000e-01 2.98999995e-011.48399997e+00 4.72900009e+00 3.24499989e+00 5.01700020e+005.61499977e+00 2.19820007e+02 2.14084000e+02 1.06429996e+011.06540003e+01 2.18882996e+02 3.25600004e+00 1.98300004e+002.28000000e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][1.19599998e+00 1.18499994e+00 5.87000012e-01 6.77799988e+005.02799988e+00 2.13882996e+02 2.20891006e+02 2.19294998e+022.17084000e+02 2.17460007e+02 2.21787003e+02 2.14869003e+022.17048004e+02 2.17300003e+02 3.26699996e+00 5.87000012e-010.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.09299994e+00 1.19599998e+00 4.69000006e+00 1.76100004e+003.85400009e+00 6.22399998e+00 8.58300018e+00 9.75699997e+009.74600029e+00 9.15900040e+00 9.15900040e+00 1.06540003e+011.00889997e+01 8.01799965e+00 7.69700003e+00 8.85999978e-010.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][3.21799994e+00 1.79400003e+00 2.98999995e-01 2.07100010e+001.19599998e+00 2.38100004e+00 4.73999977e+00 7.68599987e+003.55500007e+00 3.84299994e+00 2.65799999e+00 2.95700002e+005.92500019e+00 2.68000007e+00 8.97000015e-01 2.98999995e-010.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00][2.61999989e+00 1.19599998e+00 2.98999995e-01 0.00000000e+007.54999995e-01 8.97000015e-01 4.15299988e+00 1.19599998e+002.95700002e+00 8.85999978e-01 2.28000000e-01 2.28000000e-013.00000000e+00 5.27000010e-01 7.54999995e-01 3.16799998e+002.28000000e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][5.87000012e-01 3.05999994e+00 5.87000012e-01 2.21700001e+009.29000020e-01 5.27000010e-01 2.43499994e+00 2.61999989e+008.97000015e-01 5.87000012e-01 7.05999994e+00 5.87000012e-011.28799999e+00 2.28000000e-01 2.69499993e+00 8.69000018e-012.28000000e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00][1.76100004e+00 1.76100004e+00 1.76100004e+00 2.69000006e+005.87000012e-01 3.22799993e+00 2.66300011e+00 2.54900002e+001.30999994e+00 5.87000012e-01 5.87000012e-01 2.94600010e+005.87000012e-01 9.29000020e-01 1.09700000e+00 1.09700000e+005.27000010e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00]]<matplotlib.collections.QuadMesh at 0x263b7175390>
附件:
汉.jpg
四色.jpg
