Skip Cells Not Used in a Step
In their current form, our
clean_data steps both contain a
number of lines of code that generate diagnostic output or do other work that
is not core to the work of those steps. For pipeline runs, we don’t want to
include this code, because it will add unnecessary compute cycles and
We can keep this code in our notebook where it will be of use as we continue to develop our models, but explicitly exclude it from pipeline steps. To do this, we need to make sure that all the statements to be excluded are in separate cells from the code that is core to pipeline steps. Then, we can apply the Skip Cell annotation to these cells.
Please follow along in your own copy of our notebook as we annotate skip cells.
read_data step, we’ll want to skip the calls to
df.auto.describe(). To do this, edit each cell and select Skip Cell from the
Cell type pull down menu.
When you’ve done this successfully, your
read_data step will look something
like the figure below.