Inconsistent solution keys across `variable_ne_branch_*` methods

[matteorossini]

variable_ne_branch_power_real

variable_ne_branch_power_real(pm::AbstractPowerModel; nw, bounded, report) defines :pf and :pt as solution keys when assigning variables to solution dict:

PowerModels.jl/src/core/variable.jl

Line 1204 in c9544b5

report && sol_component_value_edge(pm, nw, :ne_branch, :pf, :pt, ref(pm, nw, :ne_arcs_from), ref(pm, nw, :ne_arcs_to), p_ne)

However, variable_ne_branch_power_real(pm::AbstractAPLossLessModels; nw, bounded, report) uses :p_ne_fr and :p_ne_to instead:

PowerModels.jl/src/form/dcp.jl

Line 283 in c9544b5

report && sol_component_value_edge(pm, nw, :ne_branch, :p_ne_fr, :p_ne_to, ref(pm, nw, :ne_arcs_from), ref(pm, nw, :ne_arcs_to), p_ne_expr)

---

variable_ne_branch_power_imaginary

Similarly, variable_ne_branch_power_imaginary(pm::AbstractPowerModel; nw, bounded, report) uses :qf and :qt, whereas variable_ne_branch_power_imaginary(pm::AbstractActivePowerModel; nw, bounded, report) uses :q_ne_fr and :q_ne_to:

PowerModels.jl/src/form/apo.jl

Lines 32 to 33 in c9544b5

	report && _IM.sol_component_fixed(pm, pm_it_sym, nw, :ne_branch, :q_ne_fr, ids(pm, nw, :ne_branch), NaN)
	report && _IM.sol_component_fixed(pm, pm_it_sym, nw, :ne_branch, :q_ne_to, ids(pm, nw, :ne_branch), NaN)

---

Other functions

I also performed a similar search in other variable_ne_* functions: it seems that variable_ne_branch_cosine and variable_ne_branch_voltage_product_angle are the only ones that use :*_ne* in solution keys (:cs_ne and :td_ne, respectively). Perhaps those keys could be changed into :cs and :td for consistency.

[ccoffrin]

I see the issue. Working on a fix.