Skip to content

Update docs regarding Decimal scientific notation #377

Open

Description

@Mark90
  • Include a link to the documentation section or the example

https://github.com/oracle/python-cx_Oracle/blob/master/doc/src/user_guide/sql_execution.rst#fetched-number-precision

  • Describe the confusion

Python's decimal.Decimal defaults to scientific notation for representing very small float values. Since the docs suggest using Decimal for retaining float precision, I think it makes sense to also mention this aspect, because depending on the developer's usecase this may or may not be a problem.

Python 3.7.4 (default, Jul  9 2019, 18:13:23)
[Clang 10.0.1 (clang-1001.0.46.4)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> from decimal import Decimal
>>> print(Decimal('0.000001'))
0.000001
>>> print(Decimal('0.0000001'))
1E-7
>>> print("{:f}".format(Decimal('0.0000001')))
0.0000001
  • Suggest changes that would help

Add a code snippet and explanation similar to the above to the corresponding documentation section.

Activity

  1. Mark90 commented on Dec 2, 2019

    @Mark90
    Author

    I don't know why the Bug label was added, it's not a bug. It should be labeled in relation to Documentation Improvements.

  2. anthony-tuininga commented on Dec 2, 2019

    @anthony-tuininga
    Member

    I believe that has to do with display of the value and not specifically to do with how it is represented internally. We can add a note to the documentation, however, to clarify.

  3. Mark90 commented on Dec 2, 2019

    @Mark90
    Author

    Sorry, yes, I actually meant to refer to the string representation.

    Thanks, a note would be great!

    Bit more background. When I extract data from a database (which wasn't Oracle up till recently) e.g. to load into other systems, I rarely fiddle with how data is written out, because Python and/or the DB client takes care of it. In this case using Decimal may set the developer up for an unexpected failure - but I can also see why it's a great solution to retain precision. I just noticed there was an attempt to use Decimal as default, which was reverted for the same reason :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions