The day-by-day data behind the feed report (v4): one row per pen per day with the first feeding time, head, delivered, refusal and DMI against target; the loading deviation total per day across all loads; and the checks the pipeline ran on this report. Download either table as CSV, or the whole report as an Excel workbook (one sheet per report section). The figures are the ones the PDF report prints.
Run by the pipeline every time the report is built: the same source weights summed different ways must agree, and the backup must cover the period.
Lactating pens: each pen's first delivery of the feed day, compared with its usual 1st feeding time: the start time set for the pen on the Farms page, or else the median of its 1st feeding times over the previous 60 days (rolling, per day). Thresholds and the number of days are set per farm on the Farms page (Report v3 thresholds, 1st feeding start times).
No lactating pen was fed in this period.
Operators ranked by loading accuracy over the selected range, with what their loading deviation cost at the feed system's ingredient prices on the day. Cost of over-loading = feed loaded above the ration target: bought and fed beyond the plan. Value under-loaded = feed the cows did not get, not a saving. Accuracy = 100 - (sum of |actual - target|) / (sum of target) x 100, as on the report's operator pages. Loads with no operator in the feed system are shown as Unknown and not ranked. On this page only for now; not in the report.
No operator loads in this period.
Every ingredient line of every load, all diets and all operators, summed from the source weights per feed day. Variable ingredients and scale errors (lines above 5x target) are left out, as on the operator pages. Under-loaded value is not a saving.
One row per pen per day for the report period and the comparison period before it. Times are farm local time; feeds within the farm's feeding window are counted as one feed. An empty cell means not available, never zero.