葛瑞斯肯樂活筆記

下述是使用 Tensorflow 均勻分佈的分式:

第3行是最重要的，四個參數分別描述如下:

[100]: 100個值

minval: 最小值

maxval: 最大值

dtype: 資料型態

第5行的 uniform.eval() 會執行出下述結果:

[ 0.2527045   0.33153427  0.22798097  0.29095912  0.51582193  0.73846745
  0.78685546  0.35579515  0.47969484  0.47483838  0.39843249  0.09375143
  0.70163476  0.28395784  0.44987655  0.00491154  0.94639301  0.61235738
  0.17121935  0.1124543   0.36363184  0.94005847  0.18793511  0.39175403
  0.83550525  0.55777705  0.16905439  0.23662233  0.19996226  0.3683461
  0.99905348  0.383021    0.94683552  0.36594236  0.36235213  0.35700905
  0.18580365  0.71240723  0.88535964  0.38936961  0.99490798  0.40913558
  0.27031219  0.61285663  0.57228458  0.76658046  0.61338997  0.56038547
  0.95110488  0.75753772  0.40527105  0.91629124  0.21722078  0.08193779
  0.25574708  0.95329714  0.15695989  0.29108     0.42300844  0.76490736
  0.96349239  0.26920891  0.65021968  0.54543078  0.24076641  0.81947637
  0.48174548  0.70127785  0.12298203  0.08457685  0.59318697  0.85510254
  0.07402873  0.89230418  0.73251271  0.75928617  0.18993294  0.89641702
  0.1751982   0.04954946  0.43116486  0.25196636  0.9872067   0.93730021
  0.04746079  0.89552987  0.15257907  0.20977855  0.17267191  0.63825321
  0.48887503  0.24768364  0.93892324  0.75424159  0.16232693  0.92809057
  0.80999827  0.19139016  0.17241454  0.21213412]

第6, 7行則產生下述的圖形:

 如果需要設定初始化種子，可以直接在 tf.random_uniform() 中加入一個參數 seed = 1 (數值自訂)