Flow Control

#tf.while_loop
参数
condition参数：callable returning a boolean scalar tensor
loop_vars参数：一个list或者nested list of tensor。会被传递给cond和body。`cond` and `body` both take as many arguments as there are `loop_vars`
body参数: callable returning a (possibly nested) tuple, namedtuple or list of tensors of the same arity (length and structure) and types as `loop_vars`

返回值
The output tensors for the loop variables，当cond返回是False时while_loop返回
如果loop_vars的length是1，则返回一个tensor，如果length > 1，则返回一个list

内部机制
while_loop只调用了body和condition函数各一次，将body和condition的节点构成的子图外加一些其它的op黏在一起，形成一个graph flow that
  repeats `body` until `cond` returns false。这个graph flow只会在session里运行的时候才会循环的去执行body里的计算。
  
#tf.cond
参数
pred:一个scaler
ture_fn: pred = true 时执行这个函数并返回结果
false_fn: pred = false 时执行这个函数并返回结果
`true_fn` and `false_fn` must have the same non-zero number and type of outputs

返回值
lists of output tensors

内部机制
`cond` calls `true_fn` and `false_fn` *exactly once* (inside the
  call to `cond`, and not at all during `Session.run()`). `cond`
  stitches together the graph fragments created during the `true_fn` and
  `false_fn` calls with some additional graph nodes to ensure that the right
  branch gets executed depending on the value of `pred`.
  
#tf.where
参数
condition：一个和x有同样形状，或者第一维度相匹配的tensor
x:condition对应位置为true时返回的值
y:condition对应位置为false时返回的值

返回值
当x, y 都 non-None时，返回 A Tensor with the same type and shape as x, y
当x，y都为空时：返回 A Tensor with shape (num_true, dim_size(condition))