Carry #54-slipped rolls a logbook reads at /10 in the roll table (#71)

CLAUDE.md

@@ -295,9 +295,22 @@ height missing two digits arrives, so it is gated on the band and not on `disput
 every excluded roll-height carries the reason that actually fired — no page, rows that
 disagree, a row not read are three states, and an earlier draft folded them into one.
 Spacing separates readings ~1.6× apart and no closer, so it cannot tell ×10 from ×10.764;
-the logbook can. The control roll `bc78065` suggests #54's own factor is ×10 on pre-2000
-rolls (fly#71), and half the r ≈ 2 mass is a wrong scale, not a mislabelled lens (fly#72).
-Do not change `fly_height_slip_factor()` from this; read `inst/notes/terrain-correction.md`
+the logbook can. Half the r ≈ 2 mass is a wrong scale, not a mislabelled lens (fly#72). Read
+`inst/notes/terrain-correction.md`
+
+- **#54's slipped frames are ÷10 wherever a logbook reads them, tabled where spacing
+agrees, and ÷10.764 stays for the rest** (fly#71) — the same logbooks, read blind against the 1,589, put the crew's height
+at the catalogue's divided by 10 on every slipped roll a page covers bar a one-frame typo
+(`bc78065`, `bc78078`, `bc79027`, `bc79103`, `bcc00085`), 6.7-7.6% above ÷10.764 on each. So 6
+roll-heights (299 frames) ride `flying_height_rolls.csv` at factor **0.1**, and both tables carry a
+`tail` column. **`fly_height_slip_factor()` was deliberately not changed**: the six 2003/2005
+rolls have no logbook link (the 2003 ones carry measured per-frame heights and fit 10.764 better;
+`bcc05001` lands on round feet under ÷10 and is undecided), so they and
+three stragglers (1,290 frames) stay on it, each listed in `_excluded.csv` with a reason saying
+so. The roll-table gate in `fly_footprint()` is `tab_factor != 1`, never `> 1`: `> 1` silently
+handed a 0.1 row back to #54's repair. `bc79027` reads ÷10 but spacing rejects it, because its page
+logs 80% overlap against the ~60% window. `bcb98013` frame 52 is a leading-digit typo that no
+factor recovers. Neither case moved the rule
 
 ## Gotchas
 

R/fly_footprint.R

@@ -199,6 +199,17 @@ fly_dem_coverage_min <- function() 0.95
 # Dividing by it puts all 1,589 frames between 0.80 and 1.32 of the height their own scale
 # implies. Reading the value as plain feet instead leaves them at 3.0 to 4.7 times it, and
 # not one of the 1,589 inside the band.
+#
+# That the band accepts it does not make it the defect on every roll (fly#71). x10 and
+# x10.764 are 7.6% apart and both land all 1,589 inside the band, so the terrain cannot
+# choose between them. The logbooks can: on five of the six rolls before 2003 whose
+# slipped frames a page covers, the crew's height is the catalogue's divided by 10, to 0.9%
+# (the sixth is a one-frame typo no factor fits), and the roll-heights that also pass the
+# spacing check are carried in `flying_height_rolls.csv`, which is consulted first. What
+# reaches this factor is the 2003-2005 rolls, which have no logbook (the 2003 ones carry
+# measured per-frame heights and fit 10.764 better; `bcc05001` is undecided), plus 19
+# older frames the table does not reach, each listed with its reason in
+# `flying_height_rolls_excluded.csv`.
 fly_height_slip_factor <- function() 3.28084^2
 
 # How far a frame's height above ground may sit from `scale x focal_length` before the DEM
@@ -238,20 +249,24 @@ fly_flying_height_max <- function() 16000
 # spacing between frames adjacent by number, which must imply the ~60% forward overlap the
 # flight was designed to. `factor` names the defect: 10 or 100 where digits were dropped (a
 # logbook "20.0", thousands of feet, catalogued as 2,000 ft), 1 where the logbook confirms
-# the height and the SCALE is the wrong field. `height_m` is the height used — the logbook's,
+# the height and the SCALE is the wrong field, and 0.1 where the height is ten times too
+# LARGE. That last case is fly#71: rolls flown before 2003 that #54 divides by 10.764. `tail` records
+# which side of the band the roll-height came from. `height_m` is the height used — the logbook's,
 # converted, since the catalogue's rounding survives multiplying by the factor.
 #
 # Keyed on roll, height, lens and scale together, so the table reaches only the frames it
-# was measured on. Every lower-tail roll-height it does not correct is listed with its
-# reason in `flying_height_rolls_excluded.csv`: an unlisted roll is unmeasured, not clean.
+# was measured on. Every lower-tail roll-height it does not correct, and every #54-slipped one
+# it does not reach, is listed with its reason in `flying_height_rolls_excluded.csv`: an
+# unlisted roll is unmeasured, not clean.
 # Produced by `data-raw/height_calibrate-lower_tail_rolls.R`; see
 # `inst/notes/terrain-correction.md`.
 fly_height_roll_table <- function() {
   path <- system.file("extdata/flying_height_rolls.csv", package = "fly")
   if (!nzchar(path)) {
     stop("`flying_height_rolls.csv` is missing from the installed package.", call. = FALSE)
   }
-  utils::read.csv(path, colClasses = c(film_roll = "character", cause = "character"))
+  utils::read.csv(path, colClasses = c(film_roll = "character", cause = "character",
+                                       tail = "character"))
 }
 
 # The DEM-aligned grid a single footprint is counted against.
@@ -671,7 +686,9 @@ fly_is_square <- function(footprints) {
 #' inheriting whatever is wrong with it, and the catalogue's is about 10.76 times
 #' (3.28084 squared) too large on 1,589 film frames from 13 rolls flown between
 #' 1974 and 2005 — a feet-to-metres conversion applied the wrong way round, which
-#' draws a 1:35000 frame 110 km across. A film frame states its height above
+#' draws a 1:35000 frame 110 km across. On six roll-heights flown 1978-2000 the flight
+#' logbooks show the factor is exactly 10 instead (fly#71), and those frames are
+#' carried in the roll table below. A film frame states its height above
 #' ground twice, as `flying_height` minus terrain and as scale times focal length,
 #' so the two are compared, and `height_source` records the outcome:
 #'
@@ -686,9 +703,10 @@ fly_is_square <- function(footprints) {
 #'   \item{`"corrected_roll_table"`}{they disagree, and the frame sits on one of
 #'     the roll-heights in `inst/extdata/flying_height_rolls.csv`, which the
 #'     province's flight logbooks and the spacing between adjacent frames settled
-#'     (fly#60): a height with one or two digits dropped (a logbook "20.0",
-#'     thousands of feet, catalogued as 2,000 ft), or a correct height beside a
-#'     wrong `scale`. Sized from the measured height; matched on `film_roll`,
+#'     (fly#60, fly#71): a height with one or two digits dropped (a logbook "20.0",
+#'     thousands of feet, catalogued as 2,000 ft), a height recorded ten times too
+#'     large, or a correct height beside a wrong `scale`. Checked before the
+#'     10.76 slip above. Sized from the measured height; matched on `film_roll`,
 #'     `flying_height`, `focal_length` and `scale` together, so it needs a
 #'     `film_roll` column. As with the slip, `flying_height` is not overwritten}
 #'   \item{`"implausible"`}{they disagree some other way, or `flying_height` is
@@ -704,8 +722,9 @@ fly_is_square <- function(footprints) {
 #'     `"implausible"` even where `focal_length` is missing}
 #' }
 #'
-#' Measured over the whole catalogue, the correction applies to those 1,589
-#' frames and to nothing else. A digital frame's `scale` is a nominal figure, not
+#' Measured over the whole catalogue, the two corrections for a height that is too
+#' large apply to those 1,589 frames and to nothing else: 299 by the roll table,
+#' the rest by 10.76. A digital frame's `scale` is a nominal figure, not
 #' its image scale, so it is not compared: only the 16,000 m ceiling, or terrain at
 #' or above the aircraft, can refuse one. To list the frames worth a second look:
 #' `dplyr::filter(fp, height_source != "reported")`.
@@ -1089,9 +1108,10 @@ fly_footprint <- function(centroids_sf, negative_size = 9, format_size = NULL,
     # two are compared, and a frame outside `fly_height_ratio_band()` is not sized from
     # its `flying_height` as it stands. Where dividing by `fly_height_slip_factor()` brings
     # it back inside, that is the slip and the corrected height is used; measured over the
-    # whole catalogue the repair fires on the 1,589 and on nothing else. Anywhere else the
-    # two disagree and nothing here can say which is wrong, so the frame falls back to
-    # nominal scale, exactly as one with unusable metadata does.
+    # whole catalogue the repair fires on the 1,589 and on nothing else, less the 299 on
+    # 1978-2000 rolls whose logbooks say x10, which the roll table takes first (fly#71).
+    # Anywhere else the two disagree and nothing here can say which is wrong, so the frame
+    # falls back to nominal scale, exactly as one with unusable metadata does.
     #
     # Judged after the FIRST pass and before the second, because the second pass samples
     # the rectangle this height implies and that is the expensive half of the mistake.
@@ -1123,6 +1143,10 @@ fly_footprint <- function(centroids_sf, negative_size = 9, format_size = NULL,
     # inferring one. A height slip must still reconcile THIS frame's ratio, as #54's does;
     # a factor of 1 says the scale is the wrong field, so the ratio against that scale is
     # exactly what cannot be asked of it — but that height must still clear the ground.
+    # A factor below 1 is a height recorded too LARGE (fly#71: the 1978-2000 rolls #54
+    # divides by 10.764, whose logbooks read ten times less) and is held to the band like
+    # any other slip. The test is `!= 1`, not `> 1`: the gate once read `> 1`, which handed
+    # those frames silently to #54's repair.
     #
     # Not only `disputed` frames: a height with two digits dropped sits BELOW the terrain
     # (bc7280 is catalogued at 60 m), which `disputed` hands to the terrain-above-aircraft
@@ -1154,7 +1178,7 @@ fly_footprint <- function(centroids_sf, negative_size = 9, format_size = NULL,
     # is 96 m short.
     r_tabled <- (tab_height - first$elev) / nominal_agl
     tabled <- out_of_band & is.finite(tab_factor) &
-      ((tab_factor == 1 & r_reported > 0) | (tab_factor > 1 & in_band(r_tabled)))
+      ((tab_factor == 1 & r_reported > 0) | (tab_factor != 1 & in_band(r_tabled)))
 
     slipped <- disputed & !tabled & in_band(r_repaired)
     fh_used <- fh
@@ -1293,7 +1317,8 @@ fly_footprint <- function(centroids_sf, negative_size = 9, format_size = NULL,
       warning(
         sum(roll_tabled), " of ", sum(dem_eligible), " frames sit on a roll whose ",
         "`flying_height` disagreement was settled by the flight logbooks and the spacing ",
-        "between frames (`flying_height_rolls.csv`): a height with dropped digits, or a ",
+        "between frames (`flying_height_rolls.csv`): a height with dropped digits or ",
+        "recorded ten times too large, or a ",
         "correct height beside a wrong `scale`. Sized from the measured height; ",
         "`flying_height` is left as supplied and `height_agl` is the height used. ",
         "See `height_source`.",

data-raw/flying_height_logbooks.csv

@@ -192,3 +192,24 @@
 "CONTROL_bc78065__bc78065_3.jpg",TRUE,"bc78065","78065","62-270",62,270,"13.0","True Height (M'/M.S.L.)","13000","f 153 mm","153","","partial","sheet 3 of 3; Princeton Area Rivers OP 212/78; finals overwritten in red, read as 62...270 (uncertain); 80% fwd overlap"
 "CONTROL_bcc228__bcc228_1.jpg",TRUE,"bcc228","bcc228","1-166",1,166,"12.650","M'/M.S.L.","12650","f 305","305","","clear","Remark: ""Too much turbulence for this job at 7,650' with 6"" cone"""
 "CONTROL_bcc228__bcc228_2.jpg",TRUE,"bcc228","bcc228","177-218",177,218,"14.0","M'/M.S.L.","14000","f 305","305","","clear","First final 177 (possibly overwritten 187)"
+"bc78078__bc78078_6.jpg",FALSE,"bc78078","B.C. 78078","55-63",55,63,"3.1","True Height (M'/M.S.L.)","3100","f 153 mm","153","","clear","handwritten form; sheet 6 of 16; field roll NOV. 25; camera 21162; strip 1 dir 315; op 358/78, June 27/78"
+"bc78078__bc78078_6.jpg",FALSE,"bc78078","B.C. 78078","64-76",64,76,"-","True Height (M'/M.S.L.)","3100","f 153 mm","153","","clear","height column is a ditto dash under 3.1; strip 2 dir 160; final 64 written in red slightly smudged ('CA'-like) but field 64 is clear"
+"bc78078__bc78078_6.jpg",FALSE,"bc78078","B.C. 78078","77-91",77,91,"-","True Height (M'/M.S.L.)","3100","f 153 mm","153","","clear","height column is a ditto dash under 3.1; strip 3 dir 340"
+"bc79027__bc79027_4.jpg",FALSE,"bc79027","B.C. 79 027","190-204",190,204,"6.5","True Height (M'/M.S.L.)","6500","f 153 mm","153","","clear","sheet 4 of 4 (both numbers overwritten from 3 in blue); field roll OSCAR 8; camera ZE 21162; strip 1 dir 360; field numbers partly cut off at page edge; page also notes 173-175 cleared and negs 152-162 destroyed (processor breakdown); 80% fwd overlap"
+"bc79027__bc79027_4.jpg",FALSE,"bc79027","B.C. 79 027","205-214",205,214,"-","True Height (M'/M.S.L.)","6500","f 153 mm","153","","clear","height column is a ditto dash under 6.5; strip 2 dir 180"
+"bc79103__bc79103_5.jpg",FALSE,"bc79103","B.C. 79/03","294-317",294,317,"7.2","True Height (M'/M.S.L.)","7200","f 153 mm","153","","partial","sheet 5 of 5; field roll OSCAR 43; roll written '79/03' (read as 79103); height 7.2 clear; end final read 317 (red) - 3rd digit could be 7 or 1 unlikely other; field end number crossed out and overwritten (appears 124, consistent with field 101 -> final 294 offset); strip 2 dir 005, off at Miller Creek; film ran to end, spliced onto OSCAR 42"
+"bcb98013__bcb98013_1.jpg",FALSE,"bcb98013","15BCB99013","001-123",1,123,"24000","True Height","24000","CFL 152.861","153","1: 35000","clear","typed form, height in plain feet (no unit stated; 24000 m impossible); page says final roll 15BCB99013, flight B-010-E-99, date 1999-AUG-02 - differs from file prefix bcb98013 (98 vs 99); line 197 dir 090; sheet 1 of 1; camera RC20 5085"
+"bcb98013__bcb98013_1.jpg",FALSE,"bcb98013","15BCB99013","124-194",124,194,"24000","True Height","24000","CFL 152.861","153","1: 35000","clear","line 198 dir 270; off at 194 (126 deg 15 min); same roll-number discrepancy as above"
+"bcb98013__bcb98013_1.jpg",FALSE,"bcb98013","15BCB99013","195-236",195,236,"24000","True Height","24000","CFL 152.861","153","1: 35000","clear","line 198 dir 270 resumed at 195; end 236; same roll-number discrepancy as above"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","001-021",1,21,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","typed form, height in plain feet; flight C-057-FS-00, 2000-SEP-23; camera RC 10 3539; line 12 dir 270; sheet 1 of 1"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","022-045",22,45,"18000","True Height","18000","CFL 304.483","305","1: 15000","clear","line 11 dir 090"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","046-068",46,68,"18000","True Height","18000","CFL 304.483","305","1: 15000","clear","line 10 dir 270"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","069-089",69,89,"18000","True Height","18000","CFL 304.483","305","1: 15000","clear","line 9 dir 090"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","090-111",90,111,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","line 8 dir 270"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","112-135",112,135,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","line 7 dir 090"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","136-159",136,159,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","line 6 dir 270"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","160-182",160,182,"19000","True Height","19000","CFL 304.483","305","1: 15000","clear","line 5 dir 090"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","183-202",183,202,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","line 4 dir 270"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","203-218",203,218,"18500","True Height","18500","CFL 304.483","305","1: 15000","clear","line 3 dir 090"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","219-229",219,229,"19000","True Height","19000","CFL 304.483","305","1: 15000","clear","line 2 dir 270"
+"bcc00085__bcc00085_1.jpg",FALSE,"bcc00085","30BCC00085","230-236",230,236,"19000","True Height","19000","CFL 304.483","305","1: 15000","clear","line 1 dir 090"