After a fairly long debugging session with a friendly LLM agent we have discovered that a subset of 32 measurements for pATM_au, tp21_au, pChk2_au and pDNAPK_au is off by a factor of 1000. Not all of them, but a seemingly random subset of 32. Easy to spot check by searching for "404.0" (line 1516) or "654.0" (line 1583) which will reveal measurements that are completely off scale.
After a fairly long debugging session with a friendly LLM agent we have discovered that a subset of 32 measurements for pATM_au, tp21_au, pChk2_au and pDNAPK_au is off by a factor of 1000. Not all of them, but a seemingly random subset of 32. Easy to spot check by searching for "404.0" (line 1516) or "654.0" (line 1583) which will reveal measurements that are completely off scale.