TPC: validity timestamp for time slot residual gain calibration#8523
TPC: validity timestamp for time slot residual gain calibration#8523badarots wants to merge 1 commit into
Conversation
a04ff14 to
d7fe19d
Compare
wiechula
left a comment
There was a problem hiding this comment.
Just a minor suggestion.
There was a problem hiding this comment.
would either
| static const double TFlength = 1e-3 * o2::raw::HBFUtils::Instance().getNOrbitsPerTF() * o2::constants::lhc::LHCOrbitMUS; // in ms | |
| return static_cast<uint64_t>(processing_helpers::getCreationTime(pc) / TFlength); | |
| static const auto TFlength = static_cast<uint64_t>(1e-3 * o2::raw::HBFUtils::Instance().getNOrbitsPerTF() * o2::constants::lhc::LHCOrbitMUS); // in ms | |
| return processing_helpers::getCreationTime(pc) / TFlength; |
or
| static const double TFlength = 1e-3 * o2::raw::HBFUtils::Instance().getNOrbitsPerTF() * o2::constants::lhc::LHCOrbitMUS; // in ms | |
| return static_cast<uint64_t>(processing_helpers::getCreationTime(pc) / TFlength); | |
| static const double TFlength = 1e-3 * o2::raw::HBFUtils::Instance().getNOrbitsPerTF() * o2::constants::lhc::LHCOrbitMUS; // in ms | |
| return static_cast<uint64_t>(double(processing_helpers::getCreationTime(pc)) / TFlength); |
make sense?
There was a problem hiding this comment.
It makes, I would go for option 2. Option 1 will truncate the TFlengh from 11.4 ms to 11 ms.
There was a problem hiding this comment.
Does this way of setting the validity time make sense for you? The drawback is that the validity time will not be exactly equal to the creation time, it will be the previous timestamp (in ms) multiple of the time frame interval... this is the same thing LHCClockCalibrator does.
The other option I could think of was to pass the time stamp as the TF identifier, which would set the validity to the exact creation timestamp, but this would change the semantics of CLI arguments like --tfs-per-slot to milliseconds per slot.
Also, better default values for fit selection
|
This PR did not have any update in the last 30 days. Is it still needed? Unless further action in will be closed in 5 days. |
Also, better default values for fit selection