Hi Rahul and Sanjay, I worked through the reconstruction and prepared a concrete correction, with the before/after file and reproducible tests: https://github.com/batpad/goamines/issues/1
Two date boundaries affect the result: movements on 31 March 2024 were being added to that day's closing-stock baseline, while timestamped movements on 31 March 2025 were excluded by the upper-bound comparison. The patch starts after the opening closing-stock day and includes the whole final day. On the downloaded public database, four reconstructed balances change. All 166 patched reconstructions match an independent calendar-date calculation. The issue contains the four-row report and a free download with the patch, ten boundary tests, full receipt and instructions. The test runs in memory and leaves the source database unchanged. This uses an end-of-day closing-stock interpretation. It does not resolve missing baselines or source completeness, and negative values alone do not establish physical shortage or wrongdoing. The royalty stream still has no 2024 baseline. Thank you for making the data and code available — happy to discuss the cutoff convention or adapt the patch. Bryan / CivicDataForge -- Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org --- You received this message because you are subscribed to the Google Groups "datameet" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/datameet/1d10abff-c260-47c2-b0ca-4051d7e61345n%40googlegroups.com.
