support coercing iterables with nonuniform dimensionally equivalent units to a single unyt_array

[ngoldbaum]

Fixes #210.

See the test I modified for a usage example.

[neutrinoceros]

Here are a couple suggestions. Otherwise looks great to me !

[neutrinoceros]

This is my one important suggestion here, see https://www.python.org/dev/peps/pep-3134/

Suggested change

	except UnitConversionError:
	raise IterableUnitCoercionError(str(input_object))
	except UnitConversionError as exc:
	raise IterableUnitCoercionError(str(input_object)) from exc

[ngoldbaum]

I don’t see how this is any better. It would help in the future if you would explicitly explain why a change would help instead of suggesting that someone read a long technical document to explain why.

[neutrinoceros]

Sure thing. This produce a different traceback.

with a bare raise

try:
    1/0
except Exception:
    raise ValueError

one gets

Traceback (most recent call last):
  File "/Users/robcleme/dev/yt-project/yt/t..py", line 2, in <module>
    1/0
ZeroDivisionError: division by zero

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/Users/robcleme/dev/yt-project/yt/t..py", line 4, in <module>
    raise ValueError
ValueError

This reads as if something unexpected happened in the except block, but it's not the case: we're raising a different error very much on purpose.

with a raise from

try:
    1/0
except Exception as exc:
    raise ValueError from exc

Traceback (most recent call last):
  File "/Users/robcleme/dev/yt-project/yt/t..py", line 2, in <module>
    1/0
ZeroDivisionError: division by zero

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/Users/robcleme/dev/yt-project/yt/t..py", line 4, in <module>
    raise ValueError from exc
ValueError

This is preferable because the error message conveys the designed intention better. This is exactly the situation raise from was intended for

[ngoldbaum]

I don’t think the second version is any more helpful, sorry. It’s fine as-is.

[neutrinoceros]

nice !

[neutrinoceros]

Suggested change

	from unyt import cm, m, g
	from unyt import cm, km, m, g

[neutrinoceros]

Seems a tiny bit more fragile (in a good way) like this

Suggested change

	b = [1 * cm, 2 * m, 3 * cm]
	b = [1 * cm, 2 * m, 3 * km]

[ngoldbaum]

Your suggested change makes zero difference to the actual thing we’re testing, no.

[neutrinoceros]

Suggested change

	assert_equal(a + b, unyt_array([2, 202, 6], "cm"))
	assert_equal(b + a, unyt_array([2, 202, 6], "cm"))
	assert_equal(a + b, unyt_array([2, 202, 300_003], "cm"))
	assert_equal(b + a, unyt_array([2, 202, 300_003], "cm"))