Branch into itself: core implementation#1419
Conversation
figueroa1395
left a comment
figueroa1395
left a comment
There was a problem hiding this comment.
I'm almost done, but checking out the topology tests is not an easy task. I will finish tomorrow with the last part, but so far, so good.
In addition, I will also dig into the SE to double check again if nothing breaks there.
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
mgovers
force-pushed
the
pgm/feature/lines-into-itself-core
branch
from
f832760 to
363a965
Compare
figueroa1395
left a comment
figueroa1395
left a comment
There was a problem hiding this comment.
since both the docs and the data validator at this point in time still reject branches into itself, the behavior is still unspecified (subcategory of undefined behavior) and, therefore, such a bug should not be considered to affect users, because it is the user's responsibility to not invoke undefined behavior.
I agree, but lets make sure we all agree on this so we don't have to circle back later. We also may want to send a deprecation notice via the mailing list/community meeting or similar.
On the other hand, I scanned again NRSE and I didn't see any problem thus far. There isn't any explicit check against having self connecting branches + the math seems to hold as long as the ybus entries are fine, which seems to be the case as well.
Finally, I noticed that we may have an issue with observability, as I don't think it will recognize "flow" sensors on a self connecting branch since this will only appear in the y-bus diagonal, whereas the check looks for the upper diagonal parts. This needs to be further checked, but something to keep and potentially tackle in a follow up.
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com> Co-authored-by: Santiago Figueroa Manrique <figueroa1395@gmail.com> Signed-off-by: Martijn Govers <martygovers@hotmail.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
…o itself, one voltage sensor and one power sensor on the branch Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
mgovers
force-pushed
the
pgm/feature/lines-into-itself-core
branch
from
3b01b4f to
0224416
Compare
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
…ranking Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
Signed-off-by: Martijn Govers <Martijn.Govers@Alliander.com>
mgovers
force-pushed
the
pgm/feature/lines-into-itself-core
branch
from
e51c729 to
45a257d
Compare
|
3 participants
Part of #1418
Contributes towards #35
NOTE: this only affects the core. No validation cases are added in this PR. While it is of course possible that there are bugs introduced that will be caught by such validation cases, since both the docs and the data validator at this point in time still reject branches into itself, the behavior is still unspecified (subcategory of undefined behavior) and, therefore, such a bug should not be considered to affect users, because it is the user's responsibility to not invoke undefined behavior.