basic question on how "import" works
On 19Jan2019 11:42, joseph pareti <joepareti54 at gmail.com> wrote:
>[*u23885 at c009 3_NeuralNetworks]$ cat foo.py*
>from __future__ import print_function
># Import MNIST data
>from tensorflow.examples.tutorials.mnist import input_data
>mnist = input_data.read_data_sets("/tmp/data/", one_hot=True)
>[u23885 at c009 3_NeuralNetworks]$
>when the above code is executed, equivalent of
>*$ python foo.py*
>the output is as follows:
>Successfully downloaded train-images-idx3-ubyte.gz 9912422 bytes.
>Successfully downloaded train-labels-idx1-ubyte.gz 28881 bytes.
>Successfully downloaded t10k-images-idx3-ubyte.gz 1648877 bytes.
>Successfully downloaded t10k-labels-idx1-ubyte.gz 4542 bytes.
>My questions are:
> 1. where are the write / print statements that display *Successfully
> downloaded, extracting ...*
I would expect these to be in the tensorflow package. Find out where
that is installed, then grep the source for "Successfully".
It is not uncommon for programmes to set their logging threshlds to emit
"INFO" level logs if the output is a terminal, and possibly only
"WARNING" or higher otherwise, so as to provide progress reporting to a
person but less noise in a batch environment. So you may just be seeing
info messages reporting progress. And it is likely that tere will be
some mechanism to tune how noisy that it.
> 2. I don't see any such files in /tmp/data on my system, why?
These are probably temporary scratch files. The module will be fetching
data files into the "/tmp/data" directory (which you have specified) and
unpacking and importing their data. Once that is done the file is
If you cd into that directory in another terminal and run "ls" commands
regularly _during_ the import phase, you should see the filesappear and
disappear as they are fetched, processed and discarded.
Cameron Simpson <cs at cskk.id.au>