1

I'm reading the textbook "Building Machine Learning Pipelines: Automating Model Life Cycles with TensorFlow" textbook and one example shows you how to read CSV file and convert to tf.example data structure. However I'm really confused as to what they're doing with the directories on the line:

data_dir = os.path.join(os.pardir, "data")
examples = external_input(os.path.join(base_dir, data_dir))

Even the official Tensorflow exampleGen page shows:

examples = csv_input(os.path.join(base_dir, 'data/simple'))
example_gen = CsvExampleGen(input=examples)

I just don't understand what this 'data' directory refers to? I have a jupyter notebook and my csv file called "student_scores" in the the same directory - so how would I used CsvExampleGen to ingest the data and convert to tf.example?

I've attatched both the example in the textbook and my error message as well. If anyone can help me out that would be a huge help. Mainly i'm trying to figure out how to read a local csv file using CsvExampleGen. Thanks!

Book example

Error message

cooldecola
  • 118
  • 7

1 Answers1

0

Instead of

context.run(example_gen)

try running:

from invoke import task

@task
def build(c):
    c.run(example_gen)

and make a new cell:

!invoke build
DapperDuck
  • 2,728
  • 1
  • 9
  • 21