How to use tensorflow api

tensorflow 1.2 will be updated continuously
Everything was implemented on jupyter notebook.
You can see these also with http://nbviewer.jupyter.org/

The apis implemented in here

    Summarized with series of implementation in dir **tensorflow 1.0**
    -tf.gradients
    -tf.stop_gradient
    -tf.get_variable v.s. tf.variable
    -tf.control_dependencies
    -tf.train.exponential_decay() and global_step
    -tf.train.ExponentialMovingAverage
    -tf.add
    -tf.add_n
    -tf.scatter_add
    -tf.matmul v.s. tf.multiply
    -tf.multiply
    -tf.div
    -tf.exp
    -tf.sqrt
    -tf.one_hot
    -tf.nn.dropout
    -tf.nn.xw_plus_b
    -tf.nn.embedding_lookup
    -tf.group
    -tf.reduce_sum
    -tf.norm
    -tf.cast
    -tf.argmax
    -tf.ones
    -tf.random_uniform
    -tf.sign
    -tf.floor
    -tf.nn.moments
    -tf.nn.top_k
    -tf.size, tf.rank
    -tf.range, tf.size, tf.gather
    -tf.assign
    -tf.get_default_graph
    -tf.is_inf, tf.is_finite, tf.is_nan
    -tf.identity
    -tf.cond
    -tf.clip_by_value()
    -tf.nn.conv2d_transpose
    -tf.nn.conv2d
    -tf.concat
    -tf.nn.softmax
    -tf.nn.softmax_cross_entropy_with_logits
    -tf.nn.sparse_softmax_cross_entropy_with_logits
    -tf.nn.sigmoid_cross_entroy_with_logits
    -normalize with tf.reduce_mean
    -get_shape(), get_shape().as_list(), get_shape().ndims
    -tf.shape() v.s get_shape()
    -tf.tile()
    -tf.expand_dims()
    -tf.squeeze
    -tf.stack, tf.unstack, tf.split
    -tf.split