`sum(by=)` and `at` each send a diagonal read to the other, and both refuse it

[FBumann]

Prompt: "Let’s add one parent issue for caliope and file focused issues for individual topics or link existing ones."

Note

The following content was generated by AI.

What happened?

A read at the label a relation maps a carried key to is refused by sum(by=), and the error tells you to write at. at then refuses the same read, and its error tells you to move the factor or read onto another dimension. An error must name a rewrite that loads.

Reproduction

import mathspec as ms

base = {'dimensions': {'techs': {}, 'nodes': {}},
        'relations': {'link_from': {'key': 'techs', 'values': 'nodes'}},
        'variables': {'flow_cap': {'dims': ['techs', 'nodes'], 'bounds': {'lower': 0}}},
        'objective': {'expression': 'sum(sum(flow_cap, over=techs), over=nodes)'}}

for e in ['sum(flow_cap, by=link_from, over=nodes, into=techs) >= 1',
          'at(flow_cap, by=link_from, over=nodes, into=techs) >= 1']:
    ms.to_spec({**base, 'constraints': {'c': {'dims': ['techs'], 'expression': e}}})

The whole ask is in #771. That issue proposes to let at read this. If #771 is declined, the fix here is to change both messages so that they name what loads (a relation over the pair, and a mask). If #771 lands, the sum(by=) message becomes true.

Which operating systems have you used?

• Linux

Version

main at 5fce82b.

Relevant log output

Constraint 'c': sum(by=link_from): this sum lands on the key ['techs'], so each coordinate has one term and nothing is added up — that is a read, which is at()'s. Write at(..., by=link_from, over=['nodes'], into=['techs']), or sum toward a value column.

Constraint 'c': at(by=link_from) lands on ['techs'], which the expression already carries.
A call brings the dims it lands on, so that reading it tells you what it adds. Move the factor carrying ['techs'] outside the operator, or read to a column over another dimension.