The margin threshold now sits where the model calls a direction as often as a
direction actually occurs. Nothing else.
WHY THE OLD OBJECTIVE HAD TO GO. It maximised `coverage x (precision -
breakEven)`, and this function's own comments were already the case against
it: over 98 consecutive fits of the shipped SP500 H4 model, correlation
between the chosen threshold and the win rate at it was -0.056, while the
era-to-era spread of that win rate (1.32pp) matched its own binomial SE
(1.25pp) to within 0.07pp. The margin does not rank trades. So the argmax
returned whichever of ~37 bins drew the luckiest sample, and the threshold
teleported 0.42 -> 0.04 -> 0.74 in three eras.
The response at the time was to build a null-of-the-maximum gate, an
effective-sample SE and a parsimony fallback to hold the noise down. All of
that is gone now, because fitting on calibration removes the problem instead
of bounding it: coverage is a ratio against a fixed denominator so it is well
determined at every bin, the target is a measured label rate rather than an
outcome, and nothing is maximised over a noisy curve so there is no best-of-N
to correct for. Net 174 lines out, 62 in.
It deliberately does not chase edge. It cannot - at ~0 measured edge no
operating point has more of it, and pretending otherwise is what produced a
threshold of 0.96 that still passed 60% of bars while the model called a
direction ~10x too often. The edge at the chosen point is still REPORTED,
just no longer what chooses it.
THREE READINGS OF ONE QUANTITY, AND THEY DISAGREE. "How often does a direction
occur" is measured in three places and gives ~7% (the scan's own tally), ~41%
(the era loop's counters, via this function's old coverage floor) and ~50%
(the ensemble gate's OOS base rate). They cannot all be right. Rather than
pick one silently, ScanDirectionalRatePct() and EraDirectionalRatePct() are
now named accessors, the fitter targets the SCAN - that is the tally the
operator reads, and the one "predict the labels as measured during the scan
phase" names - and the threshold line PRINTS BOTH every time it moves, so the
disagreement is on the record instead of buried in a derived floor.
The ensemble gate's own floor is deliberately NOT changed in this commit. If
the scan is right, a calibrated member covering ~7% of bars cannot clear a
12.4% floor and every model would fail the gate by construction; if the gate
is right, the scan tally is wrong. The CALIBRATION field added in 667f2bc
reports the OOS true class rates directly and settles it in one era - that
measurement comes first, and the floor follows it.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Reverts a863796 on the operator's call - "unnecessary complexity". It was
right about the mechanism and wrong about the priority: it re-cut the classes
for a case the measured verdict never reaches (SP500 H4 reads "both sides" at
the derived geometry), while the drift that IS happening affects every chart
and every era. Recoverable from a863796 if a one-sided book ever becomes real.
Two pieces of it survive, both independent of the exit idea:
The drift verdict keeps reading m_winLongCache/m_winShortCache rather than the
collapsed label pair. That line reports always-long vs always-short win rates,
which is what the win caches hold - each side scored on its own barriers,
published before the collapse. The label pair carries only the side touched
first, so it undercounted long wins by the both-won-goes-to-short share. There
are zero both-won bars at any geometry with target >= stop, so this changes no
number today; it changes the wrong number to the right one.
And the .cfg gains nothing and loses nothing: the two appended ints go away
again, and they were the last fields, so a .cfg written by yesterday's build
still reads correctly - the loader simply stops before them.
WHAT THE REVERT MAKES ROOM FOR. The operator's actual requirement is that the
model reproduce the label distribution the scan measured, and nothing in the
pipeline ties it to that. The loss trains on a rebalanced sample and the
abstain rate is owned by a margin threshold fitted on EDGE, so the call rate
and the label prior can drift arbitrarily far apart - and did, invisibly:
at era 1350 the models call Buy on 20-28% and Sell on 22-32% of bars against
a scan-measured 2.1% and 4.8%. Roughly a 10x over-call, and not one line in
the journal said so.
The era line now carries it:
CALIBRATION calls vs true rate Buy 28% vs 2% (14.0x) Sell 32% vs 5% (6.4x)
Neutral 40% vs 93% (0.4x)
Reported as a ratio because that is the readable number - 1.0x is calibrated.
This is deliberately a measurement and not yet a correction: matching the
label rate would put coverage near 7%, below the ensemble gate's own 12.4%
coverage floor, so calibration and the gate are in direct conflict and which
one yields is the operator's call, not mine.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The era-progress rate limit was a function-scope `static`:
static uint lastProgressLogTick = 0;
shouldLogProgress = (nowTick - lastProgressLogTick >= 5000);
In MQL5 that is ONE variable for the whole build, not one per object. A 5s
limit meant to keep a single model's console readable was therefore a limit
across the WHOLE ENSEMBLE, and it did not distribute fairly - it starved
whichever member finishes last, every era, deterministically.
Measured on today's run: the era barrier releases the members together and
LSTM landed 2.06s, 2.43s and 2.28s behind ConvLSTM on eras 1-3, against a 5s
window it could never reach. LSTM printed zero era lines. PAI, CONV and HYB
printed all of theirs - 3 each this run, 17 each in the 10:00 run, LSTM 0 in
both, and 0 again in the 08:32 run.
So one model in four has been training with NO per-era telemetry: no
per-class recall, no dW/W ratios, no zero-skill comparison, no deploy-bar
line. It was still doing the work - tier re-ranks, threshold fits and
exit-policy replays all appear on cadence - which is what made the hole look
like a grep that kept missing the line rather than a line that was never
written. It cost me the LSTM half of a gradient check earlier today.
Now a member, so each model rate-limits its own console output.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Found while costing whether to turn Run_Alglib_Baselines on. The suite builds
its design from BuildFeatureWindow() over DeriveHistoryBars() bars, and
neither reads any per-member state - symbol, period and feature toggles only.
So every member of an ensemble builds a byte-identical matrix and fits an
identical forest, MLP and OLS to it: four un-chunked fits over ~1000 columns
x 4000 rows, to print one answer four times. The MI suite has had a
once-per-chart gate for this exact shape since the ensemble landed; this
never got one.
Gated the same way, and here without the MI gate's caveat: that one has to
warn that the geometry scan at the end of its chain makes a DECISION, so a
donor-only run would leave the other members on a different target. This
suite decides nothing - nothing trades on it, no model is saved, and
ReportGeometryDrift assigns to nothing - so the donor's run is the complete
answer. Solo charts are unaffected.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
User request: "when an asymmetry is noticed in a market (like sp500 upward
drift) ... it does not need to predict shorts, but exit points. a sell signal
needs to be preceded by a buy so that it can say I predict we must close that
long."
Until now a LONG_ONLY verdict only BLOCKED short entries. The network went on
being trained to predict them - a third of its output capacity spent learning
an answer the direction policy guarantees it can never act on, while the
question the book actually faces (when to get out of the long) was never
asked. The two are not the same event: "a short pays" needs price to travel
the SHORT's target before the SHORT's stop, and at any geometry where reward
!= risk that is a different bar from "this long hits its stop first". The exit
is the second one.
So on a one-sided book TripleBarrierLabel re-cuts all three classes around the
only position the book can hold: Buy = it reaches its target, Sell = it
reaches its STOP first, Neutral = the horizon expired with it still open. Both
come off the allowed side's own barriers, which the walk already computed -
this reads longLost where it used to read shortWon, so it costs nothing.
Label lifespan and the timeout flag follow the allowed side too, so the
overlap correction is sized on the window this label actually spans.
DECIDED ONCE, AT ERA 0, AND PINNED. m_exitTargetSide goes in the .cfg beside
the derived geometry under the same doctrine and for the same reason: it
decides what Buy and Sell MEAN, and a target that moved mid-run would retrain
a fitted model against something it never saw. A .cfg from before this ends
early and reads 0/0 - "not decided, symmetric" - which is exactly what every
existing model was trained as, so nothing needs migrating. The weights
fingerprint keys on the INPUT only (explicit Long only / Short only); under
Intelligent the measured verdict must never reach a filename, or the model is
orphaned the moment more history downloads.
THE DRIFT VERDICT HAD TO MOVE OFF THE LABELS FIRST, and it turns out it was
measuring the wrong thing anyway. It counted m_labelCacheBuy/Sell and called
them "always-long vs always-short win rate", but the label pair is the
COLLAPSED first-touch verdict: a bar where both sides reached their target
carries only the side touched first, so long wins were undercounted by the
both-won-goes-to-short share. m_winLongCache/m_winShortCache are the actual
per-side win rates, published before the collapse, and that is what it reads
now. Necessary as well as more correct - deriving the verdict from labels the
verdict shapes is a feedback loop, since Sell-as-exit is near complementary
to Buy and would close the very gap that produced it. The gap's SE now leans
conservative rather than anti-conservative for the same reason.
LIVE. The retargeted class is wired to close the position, or training it
would be pointless: CheckClosePosition's "never vote-exit a certified
position" rule keeps governing symmetric books and gains a one-sided
exception, and the replay reads the identical rule through one
LiveVoteExitThreshold() so certified and traded cannot describe different
policies. Armed only when the operator picks a close threshold
(Signal_ThresholdClose ships Disabled) AND the model's own pin says its
blocked-side class means "close" - a model trained symmetric never fires it,
whatever the verdict has since become. This does trade a different game from
the one the win-rate certificate grades; the era's EXIT-POLICY REPLAY line
already reports expectancy in R for exactly this case and says so in words.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Found auditing pointer discipline, per the standing rule that CheckPointer
comes before every dereference.
THE LEAKS. CNeuronLSTM::feedForward allocated forget_gate, input_gate,
output_gate and new_content on the heap and deleted them only on the success
path. Eight error returns sit between the first allocation and that delete,
and every one of them abandoned whatever had been built so far. calcHidden-
Gradients was the same shape with fourteen returns past MemoryGradient. This
is the CPU path, which is the only path this machine has - no OpenCL, no
DirectML - so it ran on every era of every LSTM and CONVLSTM member.
Fixed by construction rather than by adding deletes: none of the five buffers
escapes its function, so each is now an automatic object. The return itself
destroys them, which means the leak cannot come back the next time someone
adds an error path - which is exactly how it got here.
CalculateGate had to change shape for that: it now fills a caller-supplied
CArrayDouble and answers bool, instead of handing back an object each caller
was responsible for deleting on its own error paths and none of them did. It
also allocated BEFORE testing `gate`, leaking on that very check, and never
tested `sequence` at all before dereferencing it. Both arguments are checked
first now. Protected virtual with three call sites, all in this file - no
public API moves.
THE UNCHECKED DEREFERENCES. The input-gradient loop did four rounds of
`temp = SomeGate.At(i); con = temp.getConnections().At(n); value +=
temp.getGradient() * con.weight` with no check on either pointer, and At()
answers NULL for an out-of-range index rather than failing loudly. The four
copies are now one AccumulateGateInputGradient() that checks the layer, the
neuron and the connection. The line above them read `temp.getConnections()`
off whatever the previous loop happened to leave in `temp` - NULL if
OutputLayer was empty - and is now checked too.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two UX changes the operator asked for.
THRESHOLDS. Signal_ThresholdOpen/Close were raw ints with the legal range
written in the label ("[0...100, 101 = never]") - the one input style this
codebase converted away from everywhere else. Open now takes the existing
PERCENTAGE_PRESETS, whose comment already declared itself to be "Signal_
ThresholdOpen's scale" but was never wired to it; Close takes a new
SIGNAL_CLOSE_PRESETS carrying the same rungs plus CLOSE_DISABLED = 101, which
is why it cannot just reuse the other enum. Member names are prefixed because
MQL5 enum members share ONE flat namespace - a bare PCT_25 in the second enum
would silently resolve to the first one's, warning only. Values are unchanged,
so existing .set files keep their settings. Both call sites now cast
explicitly at the CExpertSignal boundary rather than leaning on an implicit
enum-to-int conversion that only warns.
ARROWS. 2026-08-19 replaced the low/high arrows WITH trigger-price lines; that
was a swap where it should have been an addition, and it cost the zoomed-out
view. A mark is now both objects: the line is the precise entry/exit level,
the arrow off the candle's extreme is the finder that says there is something
here to zoom into. The arrow's name is the line's plus a suffix, so it stays
inside SIG_ARROW_PREFIX and every prefix-scoped purge already reaches it.
The two type-filtered sweeps had to widen or they would clear one half and
leave the other: the Hide/Show visibility loop and the pre-rescan scoped
delete both walked OBJ_TREND only. Both are typed-blind and prefix-scoped now
- the same widening this file's 2026-08-09 note describes, for the same reason
it gives. Deletes go through one WarriorDeleteSignalMark() so an arrow cannot
outlive the line it belongs to, and the sidecar deliberately still records one
row per mark off the line (the half carrying the price), with the restore
redrawing the pair.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The detector claimed "-1 means an INVALID HANDLE, 0 means created-but-never-
calculated". This run disproved it with our own instrumentation: the repair
line prints only when Create() RETURNED TRUE, and the depth it read
microseconds later was
BEFORE: MA=-1(h13) | AFTER: MA=-1(h13)
A freshly created, valid handle read -1 - the value the model says is
impossible for one. So BarsCalculated() < 0 does not mean "dead"; it also
covers "valid, not calculated yet", and the trigger cannot separate them.
Consequences, all fixed here:
- The AFTER depth was re-read synchronously, when it can only be -1 or 0, so
every repair looked like a failure and the line was unreadable either way.
It now reports the handle NUMBER across the recreate instead. A changed
number proves a new instance; SAME means MT5 handed back the same
refcounted one, so it was never dead.
- IndicatorDepthReport printed the handle number for MA alone. Every tunable
gets one now, through a single IndicatorDepthField() - nine near-identical
StringFormat calls collapse to one.
- The comment justifying "never release before re-creating" rested on the
claim just disproved. The decision stands, the reason is restated: given
the ambiguity, releasing is the dangerous half - a recycled number would
decrement whatever owns it now and CAUSE this outage - while re-creating a
live handle only leaks a reference on a path that fires a few times a
session.
The before-handle is captured on its own line, never as a sibling argument to
the Init* call: MQL5 does not define argument evaluation order.
Behaviour is otherwise unchanged - same trigger, same cooldown, same
recreate. Only what gets reported changed.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
ExpertSignalAIBase.mqh 4388 -> 4256 lines. Every edit verified the same
way: non-comment lines extracted before and after and diffed, so the
change is provably comments only.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Box header per function/class, section banners, and 1-3 line notes for
variables - not the paragraphs that had accumulated. Both files verified
mechanically rather than by eye: every declaration in Inputs.mqh (159 of
them) and every non-comment line in ExpertSignalAIBase.mqh is byte-
identical before and after.
Variables/Inputs.mqh 881 -> 355 lines (704 -> 179 comment lines)
Expert/ExpertSignalAIBase.mqh 4795 -> 4388, first ~800 lines converted
What was cut is narration - the history of what a constant used to be,
paragraphs restating the line below them, and commentary about earlier
versions of the comment itself. What was kept is every constant, every
measured number, and the traps worth a warning.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Reported symptom: one member at era 17 while the rest sat at era 2, with
the combined vote never scoring. Two faults compound to produce exactly
that, and neither needs a broken model to trigger.
FIRST - BUSY WAS READ AS STUCK. BarrierEraHeartbeat() decides liveness
from one signal: has m_eraCount changed in the last 12 minutes. But
Train() returns early, before the era loop, for three ONE-TIME phases
that never touch m_eraCount - the label-cache prebuild, the pattern-DB
backfill and the OOS simulation walk - and those are precisely what a
slow topology spends its first many minutes doing. A member grinding
steadily through a prebuild therefore looked identical to a dead one and
was dropped from the barrier at startup, before it had trained a single
era. The constant's own comment states the flawed premise: "comfortably
past the slowest healthy ERA on the deepest chart" - true, and not the
question being asked. Those three branches now call NoteBarrierProgress()
and a chunk of phase work re-arms the watchdog exactly as an era does.
SECOND - EXCLUSION HAD NO BOUND. Once dropped, a member is skipped by
EnsembleMinTrainingEra(). Drop every OTHER member and that loop finds
nothing to take a minimum over, falls through to its `return m_eraCount`
fallback - the CALLER'S own era - and EnsembleEraBarrierHolds() evaluates
`era > era`, false, for everybody. The barrier silently becomes a no-op
and the fastest member runs away unbounded. EnsembleMinEraAnyMember()
now measures against every still-training member, excluded or not, and a
member may lead it by at most ENSEMBLE_MAX_ERA_LEAD eras.
The cap is deliberately a real stop rather than a warning. A
desynchronised ensemble is not a degraded one: the combined-vote score
and the joint checkpoint both require every member on the same era, so
weights trained past the cap can never be certified by any gate. The
hold reports which of the two it is, because the operator's next move
differs - an ordinary barrier hold resolves itself, a lead-cap hold names
a member that needs diagnosing and will not resolve on its own.
Not yet explained: "only one NN listened to the stop command". The panel
now dispatches down the filter tree and reports the count it reached
("training stopped (N model(s))"), so the next run answers that
definitively instead of leaving it to inference.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two arrays parallel to m_isTrainQueue existed only to serve replay: a
per-occurrence sample weight, and a "count this bar once" flag that kept
the reported IS accuracy on the natural class distribution while backProp
trained on the oversampled one. Replay was removed on 2026-07-31 - every
bar has been queued exactly once since - and the scaffolding was left
standing, provably constant:
m_isTrainQueueWeightScale[] - written 1.0 at both queue sites, swapped
through the Fisher-Yates shuffle to stay in lockstep with nothing,
read into a variable passed to backProp, whose own default is 1.0.
m_isTrainQueuePrimary[] - written true at both sites, swapped the
same way, and read as two guards that could not be false.
Also a `for(int rep = 0; rep < repCount; rep++)` around a hardcoded
repCount = 1, and both arrays preallocated at totalIter * 4 - about
1.5MB per model of always-constant data, on a six-core box that trains
four of them at once. Behaviour is unchanged by construction: every
removed read had one possible value.
m_maxClassSampleWeight goes with them - declared, initialised to 1.5 in
the constructor, read nowhere, and documented as "currently unread by
that path... kept in case a smaller, additive loss-level nudge is ever
reintroduced". That is the definition of YAGNI, and its 17-line comment
described the class-balance correction as data-level oversampling, which
has not been true since July.
Comment pass on ExpertSignalAIBase.mqh, -164 lines this session with
every constant and measured number kept. One correction worth naming:
the header carried 15 lines arguing for HARD 0/1 one-hot targets and
explaining why smoothing was no longer needed - directly above
LABEL_SMOOTH_HIGH 0.9 / LABEL_SMOOTH_LOW 0.05, which every training path
actually uses. It described the opposite of what ships.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The control panel drove training by looping g_aiSignals[] - a
hand-maintained, MAX_AI_SIGNALS-capped, AI-only registry that had
already dropped an ensemble member on the floor once (609be10). A model
missing from it still trains and still votes, it just cannot be paused,
stopped, deployed or reset, and every button label is computed from the
same short list, so the panel described one set of models while acting
on another. Classic signals could not respond to a panel action at all.
Commands now walk the signal tree CExpert already owns:
Expert.DispatchSignalCommand(cmd) -> root signal -> every filter,
recursively, returning how many actually acted.
CExpertSignalCustom carries the seam (OnSignalCommand / HasSignalTrait,
both no-ops by default), so a classic signal opts in by overriding two
methods and needs no registration and no cap. CExpertSignalAIBase
implements the training commands over its existing Pause/Stop/Deploy/
Reset methods - the behaviour is unchanged, only its reach reported.
Button labels ask the same tree via CountSignalTrait, with
SIGTRAIT_TRAINABLE as an explicit denominator: "all paused" is
meaningless without knowing how many could be paused. Pause/Stop resolve
their toggle direction ONCE in the EA and hand every model the same
plain command, instead of each re-deriving the direction from its own
local state - which is how a mixed set ends up half paused. The alerts
now report the count acted on rather than assuming it.
Two dispatch bugs found on the way, both from a database guard copied
onto event delivery: CExpertSignalCustom::OnTickHandler and
::OnChartEventHandler each skipped any filter whose GetFilterID() is
"NULL". That id is a DB folder name, and CSignalNewsFilter,
CSignalSessionFilter and CSignalRiskGuard never set one - so all three
were silently receiving neither ticks nor chart events. The guard stays
where it belongs, on the paths that write pattern tables.
ENUM_CP_ACTION moves to Enumerations\GlobalEnums.mqh (now include-
guarded) because the Expert bases have to name it and the panel is
included long after them.
The AI-only lifecycle loops - PollTraining, the weight autosave,
AltDataReload, OnDeinit's shutdown cascade - still use g_aiSignals[] and
are untouched here.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Every AI signal repeated the same five-line InitIndicators override that
did nothing but call InitNeuralNetwork. The cause was an access mismatch,
not a design: CExpertSignalCustom declares InitIndicators public, the AI
base redeclared it PROTECTED, and each subclass had to redeclare it
public to be reachable by CExpert. Worse, the base's own override does a
different job entirely - it creates the OHLC/ZigZag feature indicators -
and InitNeuralNetwork called it back scope-qualified to stop the virtual
dispatch landing in the subclass. Two jobs, one virtual name, and a
recursion trap held off by a scope qualifier.
The feature-indicator step is now InitFeatureIndicators() (protected,
non-virtual, named for what it does) and the AI base carries the single
public InitIndicators override. CONV/HYBRID/LSTM/PAI/META drop their
copies and are now purely identity plus topology, which is the classic
signal file's shape.
Comment pass on ExpertSignalAIBase.mqh, -100 lines with every constant
and every measured number kept. Three claims in the tier block were
stale and inverted - it named CalibratedConfidenceMagnitude() as the
tiering input where the code deliberately uses the RAW magnitude, and it
described the signal DB as re-ranking each tier when ApplyPatternWeight
declines the DB from the end of era 1. Also dropped a paragraph whose
subject was a previous version of the comment, and moved two notes down
onto the constants they document (CONV_COMPRESSION_DIVISOR was 16 lines
and three unrelated defines away from its own text).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
targetHour CH_23 -> CH_MARKET_CLOSE, targetMinutes CM_45 -> CM_5.
A fixed 23:45 was a guess at one broker's server offset. It is silently
wrong on any other feed, and wrong twice a year on the same one.
"Market close" resolves per day from the symbol's own session table and
backs off the minute value, so it is correct on every symbol and on both
sides of a DST switch with no number for anyone to maintain.
THIS RE-KEYS EVERY FINGERPRINT and forces a fresh run from era 0. That
is correct rather than collateral: the close-all schedule is where the
label walk stops treating a trade as holdable
(Expert\AIBase\Labels.mqh), so changing it changes every label, and
models trained against the 23:45 barrier are confounded for this one.
The CUT: field in BuildModelFingerprint() catches it automatically -
verified present rather than assumed.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The MECHANISM was already stdlib and is untouched: ThresholdOpen() ->
m_threshold_open, tested as `m_direction >= m_threshold_open` exactly as
CExpertSignal does it. What was wrong was the presentation. Both inputs
were preset ENUMS labelled "Min confidence to open/close (%)", which
names the wrong quantity - m_direction is a WEIGHTED MEAN OF PATTERN
WEIGHTS, not a probability, and nothing in this path is a confidence.
They are now plain ints named the way the MQL5 wizard names them:
input int Signal_ThresholdOpen = 25; // [0...100]
input int Signal_ThresholdClose = 101; // [0...100, 101 = never]
Values are exactly what shipped, so behaviour is unchanged. 101 rather
than the library's default of 100 for close: a weighted mean of pattern
weights cannot REACH 101, which is how the shipped config disables the
vote exit, and quietly lowering it to 100 would re-arm a live exit route
as a side effect of a naming change.
VOTE_CLOSE_PRESETS is deleted (its only user is gone). PERCENTAGE_PRESETS
stays - MinRecall genuinely is a percentage.
** ACTION NEEDED ON DEPLOYED CHARTS: the inputs are RENAMED, so saved
.set files no longer match and charts fall back to the defaults above.
Those defaults are the current shipped values, so a chart on 25/Disabled
needs nothing; a tuned one does.
Comment cleanup in the same pass, and this part was not cosmetic - three
blocks documented mechanisms that no longer exist:
- the AI early-exit route (deleted in 38a12a2) described as live and
still firing every bar;
- the m_lastNonNeutralSignal alternation gate (removed 2026-08-01)
described as consuming the AI's vote;
- 16 lines of VOTE_CLOSE_PRESETS documentation orphaned by that enum's
deletion, ending with "see that enum's note directly above" pointing
at nothing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Three ALGLIB additions, all measurement-only and all under the existing
Run_Alglib_Baselines switch.
MinBLEIC COMBINING WEIGHTS. The live ensemble weights each member by its
own pooled holdout win rate - a defensible prior, but not a fit, and
nothing has ever asked what mixture minimises error on the bars the
members disagreed about. Two individually-mediocre members wrong in
different places can beat one individually better, and a per-member win
rate cannot express that because it never looks at them jointly.
Solved on the simplex (w >= 0, sum w = 1), which is exactly what
MinBLEIC is for. Non-negative because a negative weight asserts "trade
the opposite of this member", a claim ~60 effective observations cannot
support. Least squares on the signed outcome rather than precision:
precision is a STEP function of the threshold that no gradient method
can walk, and optimising a smooth proxy for a step decision is how
c3daded put every operating point 14pp underwater - so the result is
reported in BOTH currencies, the SSE it minimised and the directional
hit rate the mixture would actually have scored against the equal mix.
If the second does not improve, the first is noise.
This needed data that did not exist: g_ensVoteSum accumulates member
contributions and the sum destroys the decomposition, while
g_ensVoteVoterMask records only WHETHER a member voted, never what.
g_ensVoteMember[] keeps them unsummed. The live arithmetic is untouched.
ENS_MAX_MEMBERS is 8 and deliberately larger than MAX_AI_SIGNALS (5):
independent caps, over-allocating is free, and matching them would make
this array silently short the day the registry grows - a cap that has
already dropped a member once without saying so.
MLPKFoldCVLBFGS. Every baseline row carries a binomial SE, which is the
sampling error of SCORING a fixed model and says nothing about how much
the FIT moves. One LBFGS run from one random start can land anywhere,
and a baseline that cleared or missed the bar on luck of initialisation
reads exactly like one that did it on merit. 3 folds, because each is a
full retrain. LBFGS not LM - LM builds a Hessian over ~7,700 weights.
CCorr ALL-LAGS PROFILE. Added BESIDE the MI lag profile, not instead:
MI catches nonlinear dependence and is the stronger negative, which is
why it settled the verdict - but its per-lag permutation null limits it
to ~20 lags. FFT correlation gets every lag in one O(n log n) pass, so
linear structure parked at lag 300 would surface for free. Different
question, not a replacement. Walks CONTIGUOUS bars, unlike everything
else in this file, because a lag index is meaningless otherwise; both
series are mean-centred first since CorrR1D is a raw sum of products;
and the max over columns x lags is judged against a Sidak family of
exactly that size, not a bare 2-sigma line.
fasttransforms.mqh needed its own include - verified that none of
ap/optimization/statistics/solvers/linalg reaches it.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Asked whether ALGLIB's perceptron should replace ours "only if it is
better". This measures that instead of assuming it, and the measurement
is worth more than the swap would have been.
The forest and the linear fit test whether the MATRIX carries direction.
Neither tests whether OUR CODE is what fails. ALGLIB's MLP does: it is a
long-standing independent implementation of roughly what our own dense
stack computes, trained on the identical rows through the identical
gate. That distinction is not hypothetical here - two silent
implementation faults have already invalidated every model-based
negative behind them (a transposed dense hidden gradient, so backprop
was never backprop; and Adam storing sqrt(v) and feeding it back as
variance, so every "Adam" run was SGD). If alglib-mlp clears a bar our
own net cannot on the same rows, the fault is in our code.
Trained BEFORE the target column is rewritten for the linear fit:
MLPCreateC1 builds a softmax classifier whose trainer wants the class
index, which is what the matrix still holds at that point. Scored by
argmax with Neutral abstaining, exactly as the forest is. Sidak family
is now 3, not 2.
Deliberately NOT a replacement, and the gap is not close: ALGLIB's MLP
is feed-forward only (no conv, no recurrence - so it could not stand in
for three of the four members at all), and it has no batch norm, no
logit-adjusted loss for the class imbalance, no era resumption, no
OpenCL or CPU-DLL backend and no .nnw. Adopting it would trade a
maintained architecture for a control.
Kept small on purpose - 8 hidden units, 1 restart, 100 iterations, decay
at the 0.001 the ALGLIB docs recommend when you have no reason to pick
another. A wide net would spend the six-core box's hours answering a
question about capacity when the question is about implementation.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
MetaEditor: "declaration of 'eta' hides global variable" (Math.mqh:792
vs Network.mqh:80). The standard library's Math\Stat\Math.mqh declares a
local `double eta` in its incomplete-gamma branch, and our bare global
of the same name is in scope there.
Same fault as the b1/b2/lr/momentum macros retired in ea2552e: a
single-token global name living in a header that library code gets
compiled beside. The library cannot move, so ours does - 112 references
across 13 files, whole-word only.
Named g_eta rather than g_learningRate to stay inside the vocabulary
already around it (ETA_DECAY_FACTOR, ETA_MIN, m_etaCeiling, etaBefore),
all of which are untouched and none of which shadow anything.
One log line said "continuing to explore without decaying eta", where
the word was prose rather than a symbol reference; that reads "the
learning rate" now instead of naming a variable at the trader.
Scanned for the next occurrence rather than waiting for it: the only
other bare lowercase globals in the tree are eaName and tableschema,
both distinctive enough not to collide with a library local.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
DeriveBarrierGeometry() reads the stop off a quantile of the adverse
excursions in the IS region ONLY - correctly, since a geometry chosen
with the holdout in view has used the holdout for selection and it stops
being a holdout. The cost of that correct choice is that nothing ever
checked whether the distribution it measured still holds on the bars the
model actually trades.
If adverse excursions run wider in the OOS window than in the IS region,
the derived stop is too tight for the market it is used in, every label
was cut on the wrong geometry, and the deploy gate certified a game the
trade is not playing - the 2026-08-09 geometry mismatch arriving through
drift rather than through a config error.
Reported in TWO currencies deliberately. A rank-test p says whether the
distributions differ; it does not say whether anyone should care. The
stop each half's own quantile would derive says exactly that, in ATR
multiples - the units the order is placed in. A significant p with both
stops on the same ladder rung is a curiosity; half an ATR of movement is
a problem whether or not it clears 0.05.
Declustered first, same as the regime test: overlapping labels are not
independent draws. Harvest guards are copied from the deriver's own so
the two describe the same sample - and the split is verified identical
(totalIter == bars - historyBars, so Train's oosCutoff and the deriver's
are the same number).
Runs before the two model fits and needs only the excursion caches: a
chart too thin to fit a forest can still have drifted.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
MQL5's MathRand() is the 15-bit MSVC LCG - 32768 distinct values and
the lattice structure that shape of generator has. Two places here
actually lean on randomness and both were hurt by it:
WEIGHT INIT. Six He/LeCun-uniform sites drew
((MathRand()+1)/32768.0 - 0.5) * 2 * scale, so a first dense layer of
~250k weights had only 32768 possible values and thousands of
connections started byte-identical. Breaking that symmetry is the whole
job of random init.
SHUFFLING. ShuffleRandomIndex() already had to splice TWO MathRand()
draws to reach 30 bits, and its own comment documented the residual
modulo bias it still carried. HQRndUniformI() is rejection-sampled and
exactly uniform, so the splice and the bias note both go.
CHighQualityRand is L'Ecuyer's combined multiplicative congruential
generator - two differenced streams, 31-bit output, period ~2.3e18 -
and it ships with the terminal.
AND A BUG THE MIGRATION EXPOSED. The three MathSrand(GetTickCount())
calls sit immediately before "build a fresh topology", once per model.
GetTickCount() steps in ~15.6 ms on Windows and an ensemble builds every
member inside one OnInit, so members could be handed the SAME seed and
draw the SAME weights wherever their shapes coincide - and members that
start identical are not an ensemble. WarriorRandSeed() takes a salt (the
model id) plus a never-reset call counter, so a collision is impossible
rather than merely unlikely, while the tick keeps the run itself
genuinely unrepeatable the way those call sites asked for.
Seeds are masked positive rather than trusted: HQRndSeed computes
s % (M-1) + 1 and MQL5's % keeps the sign, so a negative seed leaves the
generator in a state its own assertions reject. GetTickCount() is a uint
and goes negative as an int after ~24 days of uptime - a fault that
would surface as "training is broken" on a long-running terminal and
nowhere else.
The indicator tuner's 52 draws move across too: its random search is
where sample quality earns its keep.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two ALGLIB diagnostics on the pass that already builds the matrix, under
the same Run_Alglib_Baselines switch, so they cost nothing by default
and share one walk of the feature pipeline.
REDUNDANCY (CBaseStat::PearsonCorrM + CPCAnalysis::PCABuildBasis).
The alt-data block already produced this finding once, by hand: 13
features x 16 window slots = 208 inputs spanning ~11.5 effective
dimensions, and the cost was not wasted parameters but a gradient
weighting bias, since batch norm rescales collinear copies without
decorrelating them. This generalises that measurement to every column
instead of the one block someone thought to check.
Run on the ANCHOR BAR'S ROW, not the full window. The window is
m_historyBars near-copies of the same columns at different lags, so
per-bar redundancy is the question worth asking - the lag structure is
what the conv/LSTM stage exists to exploit - and it keeps the eigensolve
at ~60x60 rather than ~960x960.
REGIME STABILITY (CMannWhitneyU::CMannWhitneyUTest) on the net's own
signed margin: its conviction, signed by whether the conviction paid,
older half of the OOS window against newer.
DECLUSTERED FIRST, which is the whole reason this is not a one-line
call. Triple-barrier labels overlap, so consecutive scored bars share
the price action that resolved them and are not independent draws.
Feeding all of them to a rank test yields a p computed on an n it does
not have - the same error label overlap invalidated everywhere else
here. One sample per label horizon costs most of the rows and buys a p
that means something; when too few survive, it says so and declines.
Both report and neither acts. Wiring the regime flag to position sizing
is the obvious next step and the one to take only after watching the
number: run every era on ~60 effective observations, a 0.05 test crosses
by chance regularly, and an auto-derisk on that is a random position
size generator.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Trailing, money management, settings validation, indicator setup and
timer registration each carried their own copy of the same 18-line
retry loop - a bool, a counted loop, a RandomSleep backoff, and two
Print lines - differing only in which call they made and what they
called it. Five places for the retry count, the backoff and the failure
wording to drift apart, and OnInit was 635 lines partly because of it.
RetryInitStep() is now the only copy. `what` completes both sentences
the loop printed, so the journal reads exactly as it did; `caller` is
passed in rather than read from __FUNCTION__ so the line still names
OnInit and not the helper.
The five steps become one-line wrappers because MQL5 function pointers
bind neither a method call nor an argument, and these are two of each
(Expert.ValidationSettings/InitIndicators, and the timer's interval).
That interval moves to WARRIOR_TIMER_INTERVAL_MS beside its wrapper,
taking its full rationale with it instead of leaving it stranded in the
middle of OnInit.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
InitNeuralNetwork() was 673 lines and the fingerprint assembly - the
single most audited block in the file, since its hash decides when a
trained model may be resumed and when it starts again from era 0 - sat
in the middle of it with no name of its own.
BuildModelFingerprint() is now that block, moved line for line. Its
header states the two rules the per-field notes have been repeating one
at a time for months: measured quantities never enter (the .cfg carries
those, adopt-don't-compare), and new fields append conditionally so
shipping one does not re-key models that never use the feature.
The assembly is byte-identical - verified by diffing every `fp =`/`fp +=`
line against HEAD, which differ only by the new call site. So no
existing .nnw/.cfg re-keys and nothing retrains.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The Beta-prior arithmetic that turns counts into a ranking weight was
written twice, term for term: WinRateFromCounts() for the classic
pattern ladders and RankTiersFromOos() for the AI confidence tiers.
Same formula, two transcriptions, and the same class of duplication the
binomial SE consolidation removed a few commits ago.
ShrunkRatePct() in System\BinomialStats.mqh is now the only copy. The
two call sites keep what genuinely differs - the classic path passes RAW
trade counts with a prior of MIN_TRADES_FOR_WIN_RATE, the AI path passes
OVERLAP-CORRECTED effective counts with TIER_PRIOR_EFF_N, which is far
smaller precisely because effective counts are - and that contract is
now stated once, in the function, instead of being implied by two
comments that could drift apart.
Also fixes a difference the consolidation exposed: with an empty sample
and a prior present, the posterior mean IS the prior, and returning 0
there would have handed a tier a vote weight of zero on no evidence.
The AI path could reach that (effN can round to 0 when labels overlap
heavily); the classic path cannot, since it returns NO_DATA_WIN_RATE
first.
Corrects a stale note of my own in passing: this ranking was recorded as
a "raw win rate behind a MIN_TRADES cutoff heuristic". It is not, and
has not been for some time - it is already a proper empirical-Bayes
estimator with a per-filter pooled prior. Replacing it with a
significance test, as that note implied, would have swapped the
estimator the weight needs for a gate answering a different question.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Every direction verdict so far was measured through one architecture
family, so "flat" has two readings that no topology tuning can separate:
the net is the wrong learner, or the matrix carries no directional
information.
Two learners with completely different inductive biases - Alglib's
random decision forest and an ordinary least-squares fit - now train on
the SAME feature windows (BuildFeatureWindow, the net's own function, so
there is no second feature implementation to drift), the SAME labels,
the SAME IS/OOS split with both purges, and are scored through the SAME
precision-against-always-one-direction comparison and the same Sidak
family-wise arithmetic the deploy gate uses. If both also land at
chance, the matrix is the limit.
Deliberate choices, each of which could have made the comparison a
different question wearing this one's name:
- LRBuild, not LRBuildZ: the intercept absorbs the class imbalance, and
a baseline handicapped by a forced zero intercept would flatter the
net for the wrong reason.
- Raw call counts in the SE, matching the live gate's known-permissive
test rather than correcting it here - both sides must face the same
bar.
- No threshold sweep on the linear fit: a threshold fitted on the slice
being scored is the calibration leak the purged band exists to avoid.
- Uniform stride when a cap bites, not the newest N rows, so a score
difference cannot be a regime difference. What was dropped is logged.
Ships off (Run_Alglib_Baselines = false): it is a measurement, not a
trading feature, nothing trades on the answer and no model is saved.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
First of the AI vote layers to go. CheckClosePosition had two exit routes:
the stock blended vote, and an AI-only one reading the AI members' sub-vote
undiluted. The second existed because an AI reversal averaged in with the
classic filters could be diluted below the threshold before it could close
a position.
It is gone, and with it m_lastAiVote and the aiResult/aiWeightSum pair
Direction() carried to feed it.
This CLOSES the certified-vs-traded gap rather than widening it. The
deploy gate certifies a win rate measured on hold-to-resolution outcomes,
and CheckClosePosition already gated the blended route off whenever an AI
model's derived geometry was on the order - so the AI route was the only
vote exit an AI-certified trade could take, and the exit replay existed to
reproduce it. With it removed, an AI-certified position holds to its
barrier by construction instead of by reconstruction, so Warrior_EA.mq5
now pushes ExitPolicy(0.0, true) unconditionally. Previously it forwarded
Min_Vote_Close and relied on Disabled arriving as 1.01 to switch the
simulated exit off by arithmetic - correct at the shipped default, and one
input change away from the simulation and the live path describing
different games.
Min_Vote_Close keeps its meaning for the classic route and is now
documented as inert wherever an AI certificate governs, rather than
appearing to drive an exit it can no longer reach.
Comment debt cleared while here: a tombstone block for m_ai_exit_threshold
(a member deleted 2026-08-18) still sat in the header, and four sites still
named LiveSignedConfidence's "two consumers" - it had one, the intelligent
trailing stop, since that same date.
NOT touched, and deliberately: NMS declustering is NOT a quality layer. It
gates the live signal at Inference.mqh:226 (NmsLiveAccept), and the
undeclustered population is ~8x what the EA trades. Removing it would
multiply live position count, not simplify a scoring path.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The codebase had THREE conventions for the same statistic. AltData took a
true median; the barrier horizon and the derived input window took the
upper of the two middle values; the MI terciles and the barrier stop
ladder used nearest-rank indexing. All four now go through MathMedian /
MathQuantile, which is R's type 7 and the library's one answer.
System\AltData.mqh column median -> MathMedian (exact, no change)
AIBase\Labels.mqh swing median -> MathMedian
leg-range med -> MathMedian
stop ladder -> MathQuantile, read in one call
AIBase\Topology.mqh window median -> MathMedian
AIBase\AutoTune.mqh MI terciles -> MathQuantile + MathMin/MathMax
Signals\SignalSessionFilter DST last Sunday-> CDateTime::DaysInMonth()
gaps[]/legs[] change from int to double so MathMedian can read them; the
values are bar counts either way.
VALUES MOVE. Even-sample medians shift by half a bin and the quantile
reads interpolate, so the barrier geometry and the derived input window
can land on different rungs - re-keying fingerprints and forcing a
retrain. Accepted deliberately: stdlib consistency was the ask, and three
private conventions for one statistic is what it buys out.
Two YAGNI finds fell out of the ladder rewrite. MathQuantile sorts its own
copy, so DeriveBarrierGeometry no longer sorts up[]/dn[] in place - which
means upUnsorted[], a full array copy kept only to undo that sort, is
gone. ArraySort(up) had no consumer needing order at all; it was pure
work. The library call also gets a failure guard the hand-rolled indexing
never needed but the ladder read does.
Verified while here: Math\Stat\Math.mqh's MathAbs/MathMax/MathSqrt/MathPow
and friends are ARRAY overloads, not scalar redefinitions, so pulling it
into the translation unit shadows no builtin.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The formula p(1-p)/n was transcribed nine times across six files - the two
deploy gates, the two edge floors, the collapse recall floor, the barrier
rung ladder, the inference bin SE, the pooled inverse-variance weights and
both detectability reports. System\BinomialStats.mqh now holds it once, as
free functions with no class dependency, so the god-class declaration does
not grow to host pure math.
BinomialVar(p, n) p(1-p)/n
BinomialSEPct(p, n) 100*sqrt(p(1-p)/n)
BinomialCallsForEdge(p, edge, sigmas) the same, solved for n
NormalUpperTailQ(z) Q(z), via Math\Stat\Normal.mqh
SidakFamilyP(z, N) 1-(1-Q(z))^N
Value-preserving by construction: rates go in as probabilities so no call
site gained a *100/100 round-trip, and BinomialSEPct is written through
BinomialVar so the multiply order is the one it replaced. Every degenerate
guard each site carried (p<=0, p>=1, n<=0) now lives in one place and
returns the 0 those sites already treated as "no bar to clear".
CExpertSignalAIBase::NormalUpperTail is gone; NormalUpperTailQ replaces it.
What consolidating SURFACED, and is deliberately NOT changed here: the two
Sidak selection gates compute their SE on the RAW call count, while every
other SE in the project deflates by EffectiveSampleSize() for triple-
barrier label overlap. That makes them the most permissive test in the
codebase, by ~sqrt(mean label lifespan). Correcting it tightens a live
deploy bar, which is a policy decision, not a refactor - flagged in the
code at both sites.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Six checks each fetched bid/ask and rejected a non-positive pair with their
own wording. TCLiveQuote() now owns that rule, so what counts as a usable
quote is defined once and every rejection reads the same way. The one
message that said only "no live quote for <symbol>" now names what it was
about to do, like the other five.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The gate's NormalUpperTail was a hand-rolled Abramowitz & Stegun 26.2.17
approximation. Its own comment gave the reason - "drags a chain of headers
behind it" - and that turned out to be one file: Math\Stat\Normal.mqh
includes only Math.mqh, which includes nothing. Swapped for Cody's rational
approximation in the library (~18 significant digits vs |error| < 7.5e-8).
No past verdict changes: at the z the gate operates on, the difference is
orders of magnitude below DEPLOY_FAMILY_WISE_ALPHA.
Adopting it needed the four bare macros in AI\Network.mqh gone first.
"#define b1 AdamBeta1" collides with an identifier in Math.mqh, so the
include would have macro-expanded the library's own local and failed to
compile - the same landmine that made the original author rename the
approximation's coefficients to ntB1..ntB5 rather than use the reference's
b1..b5. lr, b2 and momentum are the same class of hazard: single-token
global macros in a 52k-line codebase. All four now resolve to the input
names they always aliased, which is a pure textual identity - verified zero
bare occurrences remain.
Also:
- SelectionSort over the buffered signals was O(n^2) with an O(n^2) count of
StructToTime calls, because the comparison rebuilt both datetimes from the
six int date fields every time. Now materialises the keys once and does an
insertion sort; ArraySort cannot permute a struct array. IsEarlier goes
with it, MakeDateTime becomes SignalTime.
- Seven FileOpen sites lacked FILE_SHARE_READ|FILE_SHARE_WRITE, including
AtomicWriteBegin, which stages every model save. All 43 sites now carry
them - an exclusive open fails outright when another process holds the
path, which here has meant a silently skipped save.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
MA: CustomIndicators\ADMovingAverage is replaced by the built-in iMA (CiMA) on
both consumers - the classic vote and the NN MA input feature. This drops the
five advanced types ALMA/DEMA/ZLEMA/T3/Kalman, which have no iMA equivalent;
MA_TYPE_PRESETS is now ENUM_MA_METHOD's own codes and the tuner searches all
four. It also removes a documented failure mode: a custom indicator's depth is
bounded by TERMINAL_MAXBARS, and m_MA was the one whose feature block REJECTS
the bar on a short read - the "feature 25 fails on every bar" incident of
2026-08-17. A built-in is served at any depth.
MIGRATION. SMA moves from code 5 to 0, so persisted type codes change meaning.
SanitizeMaType() is the single validity rule; TunedPeriods records now carry a
version field and a v1 record remaps 5..8 -> 0..3, falling back to SMA for a
stored advanced type (unrecoverable - old 0..4 are indistinguishable from valid
new codes). Existing .nnw files re-key on their own, because MA_Type is hashed
into the topology fingerprint, so models retrain rather than silently running
on different MA values. EXPECT A FULL RETRAIN.
ZigZag: ADZigZag was a byte-identical rename of MetaQuotes' Examples\ZigZag -
verified by normalising identifiers and stripping comments, 233 significant
lines each with only renamed symbols differing. It now loads the stock one, so
nothing is bundled and MetaQuotes' fixes arrive without a rebuild here. Both
#resource entries are gone.
Classic_Shift: a new input, the BAR the four classic votes evaluate on (0 =
forming, 1 = last closed, default 1). One implementation on CExpertSignalCustom,
inherited by all four rather than repeated per module. Defaults to a sentinel
meaning "unset", so the AI signals and the aggregate keep the stock every_tick
rule and their feature/label alignment is untouched. The META corpus sweep still
takes precedence. CExpertBase::StartIndex turns out to be virtual, so this is a
real override, not the name-hiding the old comment claimed.
Not compiled - MetaEditor compile pending.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Verified dead by grep across all first-party sources (references/, Scripts/,
research/ excluded): EraCount, HiddenLayersCount, LstmHiddenSize, ConvFilterCount,
HistoryBars and MinTrainYear setters, PendingBatchSamples, getPrevOutIndex,
BaseCurrency, QuoteCurrency, CurrencyCount, IsLoaded, LastFiredDirection,
DBConfidence, SpecIndex, and the conv Step/WindowOut shape accessors. Every
backing member stays - each is still read internally and several are pinned by
the positional .cfg layout - so this removes surface, not behaviour.
Two comments were asserting the opposite of the code and are now true: the
"No setter: the taper's endpoints are derived" note was directly above three
setters, and the conv shape block claimed EnforceTopologyContract reads all
three accessors when CNet::FirstConvWindow only ever calls Window().
CTrailingATR::CheckTrailingStopLong/Short were byte-identical but for Bid vs Ask
and the isLong flag; both now delegate to one CheckTrailingStop body.
Deliberately NOT removed: the fractal-target branch (TrainTargetFractal,
IsFractalTarget and their label machinery). It reads as dead because the
TrainingTarget input was withdrawn, but Warrior_EA.mq5:836 documents it as a
parked option with a three-line restore path - that is a product call, not a
refactor.
Not compiled - MetaEditor compile pending.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
EstimatedInSampleBars() returned raw bars (11372 on SP500 H4) and every derived
capacity decision spent that: first-layer width, conv filters, LSTM hidden size.
But triple-barrier labels overlap - mean lifespan 9.4 bars - so the label cache line
on the same run already reports those bars are worth ~1210 independent observations.
Sizing a network against RAW bars while grading it against EFFECTIVE ones is two
subsystems disagreeing about one sample, and it disagreed in the dangerous direction
because the capacity side was the optimistic one: the warning's "roughly 1.1 weights
per training bar" is nearer 11 per independent observation.
EffectiveSampleSize() has existed since 2026-08-17 and is applied at eight sites, all
of them statistics. This adds the ninth, in the one place that decides how many
parameters get fitted. Applied inside EstimatedInSampleBars() rather than at the call
sites, because that function exists precisely so the three stages spend one budget.
SELF-ENABLING AND THEREFORE INERT WHERE IT MATTERS MOST, which is why this is two
changes and not one. MeanLabelLifespan() is 1.0 until a label cache has measured
something, so on a model's first build - before any label exists - the deflation is
correctly the identity: an unmeasured overlap must not invent a shrink. A fresh
attach constructs a fresh object, so its counters are zero too; only a mid-session
weights reset carries real evidence into a rebuild. That is deliberately safe (no
attach can now re-derive a narrower topology and discard trained weights) but it
would have left the first build - the case you most want the truth for - quoting the
flattering figure. So ReportDetectability now restates capacity against the effective
sample at the first moment L is real, for the topology already pinned. It re-sizes
nothing; it reports what was bought. Placed ABOVE that function's break-even guard on
purpose - a degenerate geometry is exactly when you want to know the net is
over-parameterised, and "it only fires for sane configs" is how the 2026-08-18
IS-error stop managed never to fire at all.
The warning also names its basis now (independent observations and L, or an explicit
"overlap NOT YET MEASURED, this is an UPPER BOUND"), so a flattering number can never
again read as a measured one.
Also factors FirstLayerFanIn() out of ComputeFirstLayerWidth so the capacity REPORT
charges for exactly what the capacity DECISION charged for - same reason
RequiredHorizonBars was factored out after the 2026-08-17 divergence - and makes
MeanLabelLifespan()/EffectiveSampleSize() const so the const budget path can call them.
Verified: no recursion (EstimatedInSampleBars -> EffectiveSampleSize ->
EstimatedInSampleBarsRaw, which computes from Bars() alone); both new StringFormat
sites hand-counted (basis 3/3 and 1/1, CAPACITY 10 specifiers / 10 arguments).
NOT COMPILED - user compiles in MetaEditor.
First clean derivation after 6f2def0 produced stop 1.22 / target 4.86 = 1:4 on
SP500 H4, and the labels came out Buy 4.3% / Sell 6.1% / Neutral 89.6% - a 20.7:1
imbalance, against 3.4:1 at 1:2 and 1.5:1 at 1:1 on the identical 10601 excursions.
That is the class-collapse regime, not a geometry.
c3daded proposed ONE leg-implied ratio per rung and left the reachability floor to
REJECT the rung, with a comment claiming the raise was "bounded" by it. Rejection is
not a bound. The 5.09*ATR median leg proposed 1:4 at every stop, all seven rungs then
missed their own floor (0.7%-9.6% against 12-20%), the no-rung-clears fallback fired
and re-proposed the same 1:4 at the tightest rung - and printed a WARNING predicting
exactly the rare positive class that followed. The system diagnosed itself correctly
and had no authority to act on it: the same shape as the scan-vs-deriver split
c3daded fixed one layer up, reintroduced one layer down.
Why the full leg overshoots: a ZigZag leg is pivot-to-pivot travel and an entry is
not a pivot - it lands inside the leg, with roughly half of it left on average.
Sizing the target at the whole median leg asks the market to deliver, from an
arbitrary bar, the entire move it usually makes between extremes.
Rather than assume the half and hard-code a factor, the raise now steps DOWN the snap
ladder (5 -> 4 -> 3 -> 2.5 -> 2) until the first-passage ladder says the target is
actually reached, stopping at the 1:2 policy floor because that is risk policy and
not a measurement. The legs still raise the ratio wherever the travel supports it;
the market decides how far. Applied in BOTH the rung loop and the fallback branch -
omitting the fallback is what actually shipped the 20.7:1 labels, since that branch
runs precisely when no rung was reachable. Rung rows now print
"[legs proposed 1:R, unreached]" so a walked-back raise is visible as one.
Verified: BarrierStepDownRr is strictly decreasing and bottoms at the floor, so both
loops terminate; rung-row StringFormat re-counted at 18 specifiers / 18 arguments.
NOT COMPILED - user compiles in MetaEditor.
Reported as "it still seems leaned towards 2x atr" after a full Delete && Reset
Weights on SP500 H4. It was not the .cfg pin, and it was not the new ratio floor
failing to take: the geometry is held in members ResetWeights never cleared, so
a reset wiped the .cfg, built a fresh net, restarted at era 0 - and then relabelled
under the PREVIOUS model's pair, before SaveTopologyConfiguration wrote that stale
pair back into the brand new .cfg. Reset, re-derive, re-pin, with the middle step
missing. No number of resets could ever have moved it.
Evidence in the 2026-08-19 journal: RESET WIPE at 16:23:38.699 (.cfg=deleted),
"rebuilt a fresh topology" at .790, and a label cache at .890 with a distribution
byte-identical to the pre-reset 2.00/2.00 one (Buy 3221 / Sell 3303 / Neutral 4837,
mean lifespan 6.4 bars) - 100 ms later, with no DeriveBarrierGeometry between them.
The member that actually blocked it is m_geometryDerivePasses.
LoadAndCompareTopologyConfiguration pins it to BARRIER_DERIVE_MAX_PASSES to block
the fixed-point iteration, which is correct for a LOAD - an existing model must
never re-derive or the target moves under fitted weights - and exactly wrong for a
RESET, which is the act of declaring there are no fitted weights left to protect.
State that is correctly sticky for one lifecycle event, silently inherited by
another; the same shape as the .cfg pin sitting beside it.
ResetWeights now returns the whole derivation to its constructor state: the three
latches, the pair, the horizon and its flags, the swing/lifespan measurements, the
scan's own outputs, and m_spreadAtr - that last one matters because the cost filter
is deliberately inert on pass 1 (m_spreadAtr <= 0.0) and an inherited spread makes
a reset model walk a different ladder than a genuinely new one. m_sl_mode/m_tp_mode
go back to the SL_Mode/TP_Mode inputs, since the adopt path overwrites them in
place with no copy of what was configured.
NOT COMPILED - user compiles in MetaEditor.
Two coupled changes, both from measurements in today's SP500 H4 log.
1. THE SCAN WAS OVERRIDING THE DERIVER ON THE WRONG OBJECTIVE.
At 13:37:47 DeriveBarrierGeometry produced stop 1.21*ATR / target 2.41*ATR - break-even
33.3%. Thirty-seven seconds later the barrier-geometry scan adopted 2:2 - break-even
50.9% - because it carried 0.0143 nats of entry-time information against the configured
pair's 0.0075, and cleared its family-wise gate. Information is not expectancy, and the
scan says so itself; nothing checked what the adoption did to the operating point. It
did this: the fitted thresholds immediately after read 38.8% win vs 50.9% break-even
(-12.2pp) and 48.3% vs 50.9% (-2.6pp), where the earlier model on this instrument at a
1:2 geometry had fitted +1.8pp. The deriver applies the ratio as user RISK POLICY; a
scan that can crown 1:1 makes two subsystems disagree about one geometry - the same
split this file already fixed once for the clamped-horizon rule. The scan now enrols and
crowns only pairings at or above the floor; sub-floor pairs are still scored and printed
(marked 'r') so the choice stays auditable. This is NOT the min-RR rule removed on
2026-08-09 - that one guarded a rejection filter that no longer exists.
2. THE RATIO IS A FLOOR, NOT A CAP (user: "the ratio of 1:2 is a minimum that I want, but
it should not cap to that if the average zigzag moves gives more room").
BARRIER_TARGET_RR -> BARRIER_TARGET_RR_MIN. ComputeBarrierHorizonBars already scanned
ZigZag pivots for leg DURATION; it now harvests leg RANGE in the same pass - two
properties of one object, so the horizon and the target describe the same legs instead
of two windows. The per-rung ratio is the floor raised toward median-leg/stop, snapped
DOWN to a coarse ladder (2.0/2.5/3.0/4.0/5.0). The ladder is coarse on purpose:
PooledGate pools only instruments whose structural break-even matches, and continuous
per-instrument ratios would never match and would silently empty the pool.
A leg is the right yardstick precisely because it owes NOTHING to the barrier - sizing a
target off travel measured over the barrier's own horizon is the circular loop that ran
EURUSD/USDCAD away to 14-31*ATR in 2026-08-07. The raise stays bounded by the three
tests already in the ladder: reachability, the horizon ceiling (first-passage time grows
with stop x target), and the cost fraction.
Consequential fixes: the reachability floor was a macro keyed to the fixed ratio and is
now BarrierMinReachPct(rr) evaluated per rung (a raised ratio has a lower break-even, so
a fixed floor would be the wrong strictness); the detectability break-even likewise;
PooledGate now writes and matches the ACTUAL ratio (TargetRR()) rather than the floor.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User report: 'eras since best' in the ensemble line is always 0 (era 147, best at era 90,
'0 eras ago'). That is a control-flow bug wearing a display symptom.
Once every member's in-sample error had plateaued, the shortcut forced the ladder to its
DEPLOY stage on EVERY era. The failed-gate branch resets the stage to 0 so the ladder can
climb again - so the shortcut raised it, the branch cleared it, forever. Three consequences,
only the first of which was visible:
- g_ensErasSinceBest was reset every era, pinning the counter at 0.
- The stage-1/2 boosted warm restarts were never reached, so the one mechanism that can
un-plateau a stuck member never ran. The models sat at a WORSE error than their best
(0.2408 -> 0.3015 on PAI) with no escape.
- Every repetition ran EnsembleSurvivesSelection against an unchanged best and incremented
the candidate-era count the family-wise correction divides by. The run spent its time
RAISING ITS OWN SIDAK BAR - the same waste as the 2026-08-18 inert IS-error stop, one
layer up, and the reason a gate that needed >47.8% saw its bar climb era after era.
Fix: the shortcut fires ONCE PER BEST-ERA (g_ensGateTestedEra, stamped before the outcome
branches because it is the re-running that inflates the family, pass or fail). A refused
gate now falls back to the normal counter-driven ladder - warm restart, anneal, then a
fresh deploy test - which is the escape the shortcut was skipping.
Also, per user: the signal marks were too small to see. Span doubled (2.6 bar widths, so
the overhang either side of the candle is ~0.8 bars) and both layers thickened - 1px dotted
was invisible on a candle chart at any realistic zoom.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Typo, and more wrong than it was: the group now holds four independent NN toggles rather
than one architecture selector.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User request: 'move from arrows on lows and highs to small horizontal lines at the actual
prices the entry/exit would trigger, just a bit larger than the candles. dark green for
buy, dark red for sell.'
Every mark is now an OBJ_TREND segment with both anchors at one price and both rays off,
spanning 1.3 bar widths, drawn at the bar's CLOSE - the price a market order actually
fires at, and the exact entry TripleBarrierLabel assumes. It used to sit on the candle's
LOW for a Buy and its HIGH for a Sell: prices the trade never touches, picked so an arrow
glyph would clear the candle. The tooltip now carries that price too.
COLOUR NOW MEANS DIRECTION AND ONLY DIRECTION on every layer (dark green / dark red).
Layer moves to width+style - the traded vote is solid and thick and drawn in front, a
single model's raw opinion is thin, dotted and behind the candles - which keeps the
distinction the old palette existed to draw (a model's opinion must never read as a trade)
while freeing colour to say one thing consistently.
Consequences handled, all of them the same 'a typed scan went blind' failure:
- SaveChartSignals filtered OBJPROP_TYPE == OBJ_ARROW and read OBJPROP_ARROWCODE. It now
filters OBJ_TREND and recovers direction from the colour. The sidecar keeps the old
217/218 numbers as its buy/sell token deliberately, so existing .arrows files still load.
- AdvanceChartSignalRestore now rebuilds through the SAME creation point the live path
uses, so a restored mark and a fresh one are identical objects.
- The rescan-scoped delete enumerated ObjectsTotal(OBJ_ARROW) - retyped, or it silently
deletes nothing.
- ApplySignalsVisibility enumerated OBJ_ARROW with NO prefix filter. Under the new type
that would have hidden and shown THE USER'S OWN trend lines on every Hide/Show click;
it is now prefix-scoped. The old type was uncommon enough on a real chart to mask the
missing check - trend lines are the most hand-drawn object there is.
- DrawObject's high/low parameters are gone (6 call sites pass m_Close instead), so no
caller can hand it a price it no longer draws at.
- Fixed a pre-existing stale comment that still described the purge sweep as OBJ_ARROW-only
three lines above the note explaining it had been widened to every type.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User request after the first ensemble run under the per-NN build: the drift verdict was
measured and printed every era while the input sat at BOTH, so it gated nothing - the
worst of both, because an authoritative-looking log line described a policy that was not
in force.
Safe as a default: the verdict fails open to BOTH (it drops a side only when the gap
clears 2 combined SEs on the overlap-deflated sample AND the weaker side sits below
cost-adjusted break-even), so an instrument with no measurable drift behaves exactly as
before. Verified it appears in no fingerprint or DB key - trade policy, not label
definition - so this changes no model identity and resets no training.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Reported by the user's MetaEditor compile of f64e0f8 (26 errors, 4 warnings). The four
warnings mattered more than the errors.
1. INTELLIGENT WAS TWO ENUMS. MONEY_MANAGEMENT_STRATEGY::INTELLIGENT (=1) is declared
BEFORE TRADING_DIRECTION::INTELLIGENT (=3) in InputEnums.mqh, so MQL5 resolved every
'tradingdirection == INTELLIGENT' to the MM member and converted it to value 1 =
TRADING_DIRECTION::LONG_ONLY. Wrong in both directions at once: selecting Intelligent
(3) matched NOTHING and silently traded both sides, while selecting Long only (1)
matched and handed the decision to the measured drift verdict - which can answer
SHORT_ONLY, so the one setting that must never go short could have. Reported by the
compiler as a WARNING only, never an error. Renamed to DIRECTION_INTELLIGENT; the
VALUE stays 3, so saved .set files are unaffected. Swept every enum in the repo for
sibling collisions (38 enums, detector validated against the pre-fix source, which it
flags): none remain.
2. g_warriorMetaGate sits above the class it points at - added the forward declaration,
the same pattern g_warriorEnsemble already uses in ExpertSignalAIBase.mqh.
3. The broker-time rename (b63e39f) never reached BufferNewTickSignal's PARAMETER or its
two call sites: the local became brokerTime, the parameter stayed gmtTime, and the
body was rewritten to read brokerTime. All five sites now agree.
4. ConfigureAISignal calls IsMetaTarget() from a free function - moved it to the public
section (identity, not an implementation seam); the other meta seams stay protected.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User design (2026-08-19): 'remove the enum menu that selects neural networks... individual
inputs for every NN just like classic signals... the META NN should be integrated into the
voting decision pipeline when enabled... as a bonus meta labelling is applied to enabled NNs.'
- AI_CHOICE is GONE (tombstoned per the stale-.set doctrine). Use_MLP/Use_CONV/Use_LSTM/
Use_CONVLSTM are ordinary bools like the classic votes; the ensemble arithmetic adapts to
any subset because the consensus divisor is the enabled capable weight. Two or more
enabled = ensemble (|ENS1 token + joint gate, exactly the old AI_HYBRID fingerprints, so
existing weight files keep loading); one = the old solo preset; none = classic-only.
- Use_MetaLabeling un-couples META from the direction NNs (the old selector made them
mutually exclusive). S3 ships: CSignalMETA::LiveMetaGate scores each vote-cleared entry
(shared window at bar 1 + proposal descriptor: side, net vote, live geometry, spread/ATR;
pattern one-hot ZEROED - ranking, not calibrated probability, documented in the body) and
vetoes below the cost-adjusted break-even. Entries only; fail-open everywhere, loudly.
- COEXISTENCE HAZARDS closed: VoteCapableWeight()=0 and ProspectiveVote()=false for the
meta target - solo-only until today, a trained META would otherwise sit in the consensus
divisor as a permanent abstainer and shrink every vote by its module weight.
- CERTIFIED == TRADED: the ensemble era verdict replays the identical veto through the same
g_warriorMetaGate pointer over its OOS fired bars (bar re-resolved from the row's own
time; fail-open counted as fires and reported: 'metaGate: N approved, M vetoed, K
unscored'). The overlay deliberately does NOT replay it (veto-filter-in-replay class,
calendar-cliff precedent) - documented at the sweep site. Solo charts' own gate does not
model the veto - the standing solo-gate caveat, documented at the input.
- DB continuity: the pattern/journal DB fingerprint's first slot was (int)AIType;
DbLegacyAiSlot() maps every legacy-expressible config to its OLD value (new 2-3 member
subsets get 100+bitmask, outside the legacy range) so no existing database re-keys.
filterID becomes the enabled roster via one EnabledNNSummary().
- HUD: the meta line shows the gate (armed/(trn), last P vs BE, ok/veto tally); the
armed/disarmed announcement fires on state change via one latch (MetaGateArmedNow), not
only when an entry happens to be proposed.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
1. Labels and the exit simulator go through the broker's stop-distance
check: risk/reward widen to SYMBOL_TRADE_STOPS_LEVEL exactly as
TCAdjustStops does at order time - the M5/tight-ATR case where live
trades ran wider geometry than training measured. Current stops level
stands in for history (like the spread); measured quantity, so it
does not key the fingerprint.
2. The Intelligent drift verdict moved into RefreshDriftVerdict(), which
RESCANS the label cache and now runs at every era end beside
RankTiersFromOos - era-cadence instead of waiting for rare full
rebuilds. Prints only on change.
3. Session filter is any-broker: sessions defined on their financial
centres' civil clocks (London 08-16 Europe/London, NY 08-17
America/New_York, Tokyo 09-18 Asia/Tokyo), converted to UTC by each
centre's own computed DST rule (EU last-Sun-Mar/Oct, US
2nd-Sun-Mar/1st-Sun-Nov), then to broker time by the MEASURED
server-vs-GMT offset (half-hour brokers included). Windows may wrap
midnight in broker time - the interval test handles it. Replaces the
EET-hardcoded anchors, which were correct on exactly one broker and
got Tokyo wrong by an hour each European summer.
4. The current-session-table-for-history caveat resolved by analysis:
the bars bound the error - a too-late assumed close meets no bars
(zero error), a too-early one truncates conservatively (<=1h, never
optimistic, cannot manufacture edge). Documented at the site.
5. The ensemble deploy gate mirrors the direction policy: blocked-side
fires are not fired bars (certified == traded), the zero-skill
reference uses only ACHIEVABLE baselines (always-short is not a
strategy a long-only book can run), and one-sidedness BY POLICY is
not degeneracy - the two-sided requirement applies only when both
sides are allowed. Sell predictions keep their other jobs (exit
triggers, consensus dilution) untouched.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User decision: "stick to the broker's time throughout the codebase and
analysis, session filter, programmed close time etc". Investigation
found the GMT choice was not just inconsistent but broken: live
journaling stamped DB rows with TimeGMT() while the online-learning
backfill stamped them with BAR time (server) - two clocks ~3h apart in
the same column. The newest-row duplicate guard compares them on one
axis, so a live row landing within the offset after a backfill row was
silently rejected as "outdated". dbVersion 3.0 -> 4.0 wipes the Signals
store: the only honest reset for a mixed-basis corpus.
- Direction()'s clock (stamps every journaled row, keys the per-second
vote window): TimeGMT -> TimeCurrent, variables renamed so the name
cannot lie about the basis.
- UpdateSignalsWeights' future-row bound: same clock as the rows.
- Session filter: broker-time anchors (London 10-18, NY 15-23:59, Tokyo
2-11). The GMT anchors were backwards for an EET-family broker - such
a broker follows European DST, so London is DST-STABLE in broker time
and moved twice a year in GMT. Tokyo drifts 1h each European summer
(no DST to track) - accepted, smallest error on offer. Also fixed:
inTimeInterval ignored its datetime parameter and called TimeGMT
fresh - a dead parameter hiding a hardwired clock.
- MetaCorpus/SignalMETA: rows pre-4.0 are GMT, broker since; the
GMT->server offset scan is KEPT because it measures rather than
assumes - it pins 0 on new corpora and still resolves old ones.
- AltDataFetch deliberately stays on GMT: FRED/COT/EIA release schedules
are external UTC-anchored events; the as-of join maps them onto server
bars downstream.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two user requests, one authority: SymbolInfoSessionTrade, read fresh on
every call so DST and per-symbol schedule changes track themselves.
- WarriorMarketOpenNow(): CheckOpenPosition refuses entries outside the
symbol's trading sessions (Sunday reopen, index CFDs' daily breaks) -
a vote can no longer fire into a closed book and collect a broker
error. ENTRIES ONLY: exits, SL/TP and the scheduled close-all stay
unguarded - closing risk must never be blocked by a session boundary.
- CH_MARKET_CLOSE = 24 (appended, .set-safe): the close-all fires
"Close-all minute" minutes before that day's LAST session close.
Friday + Market close + xxH05 = flatten 5 minutes before Friday's
actual close. Resolved identically in three places: the live executor
(CExpertCustom::OnTick), the label walk's vertical barrier
(NextScheduledCloseAll - the symbol's CURRENT table stands in for
history; MT5 keeps none, and a fixed hour is wrong by more), and the
fingerprint (the |CUT: token already carries hour=24, so switching to
the dynamic mode re-keys the model exactly like any schedule change).
Training itself is deliberately NOT gated on market hours: weekend
compute is free and labels only ever exist on real bars - what the
session table gates is order placement and, via the close-all barrier,
what the labels may count as holdable.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The schedule became part of the label's meaning (3e467f9): the same
chart trains a different target under Friday-23:45 than under
everyday-22:00. Without this token a schedule change silently resumed
weights fitted to the other target - the stale-enum-wrong-target family.
Active schedule -> "|CUT:day@hour:minute" in the fingerprint; disabled
schedule appends nothing (pre-change no-schedule models stay
byte-identical). Default-Friday charts re-key exactly once, at this
change - deliberate: their weights were trained on weekend-blind labels
and are confounded anyway.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
User report: "I exit everything on Friday close to avoid weekend swap...
if the NN training thinks I hold over the weekend it could produce
inaccurate results" - it thought exactly that. TripleBarrierLabel walked
its full horizon (64 bars, mean lifespan ~18 H4 bars ~ 3 days) straight
through the scheduled flat, scoring trades the deployed EA is guaranteed
to have closed on Friday 23:45. SQX applies this rule when building
strategies; the EA's own labels did not.
NextScheduledCloseAll() mirrors CExpertCustom::OnTick's live check
exactly (same three inputs, same -1 disabled sentinels, same
CLOSE_EVERYDAY semantics, same server clock). The walk stops at the
first bar that does not END by the cutoff - OHLC cannot order the
tradable fraction of a partial bar, and ties go to the refusal, as
everywhere in this file. An unresolved trade at the cutoff times out to
Neutral, exactly as live would flatten it. Excursions, the first-passage
ladder and the label lifespan truncate with the walk, so the DERIVED
geometry is automatically sized to the tradable window - a target the
flat rule never lets price reach stops counting as reachable.
The prebuild census now splits timeouts: "horizon too short?" vs "ended
by the scheduled close-all" - different questions, different fixes.
Schedule disabled = no cutoff, exactly like live.
Models trained under weekend-blind labels are fitted to a different
target; charts with the close-all enabled (the default) should be reset
to retrain under the honest labels.
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
SQX EdgeFinder precedent (user request): adjust for the drift instead of
fighting it. The 2026-08-19 telemetry found the models leaning SHORT
(Buy recall 21% vs Sell 40%) against a long-favored market (always-long
34.3% vs always-short 29.5% at the adopted geometry).
TRADING_DIRECTION gains INTELLIGENT = 3 (appended, explicit value,
.set-safe). It resolves at runtime from the label cache's per-side win
rates - the Buy/Sell shares ARE the win rates of taking every bar
long/short at the REAL stop/target with spread charged. A side is
dropped only when BOTH hold: the drift gap clears 2 combined SEs on the
overlap-deflated effective sample (EffectiveSampleSize - labels overlap
~18x), AND the weaker side sits below cost-adjusted break-even (a side
that still clears costs is kept; drift tilt alone is not a reason to
refuse a profitable side). Fails open to BOTH: unmeasured, tiny
effective n (<30), insignificant gap, or classic-only charts (no label
cache).
One resolution point - WarriorEffectiveDirection() - feeds all three
gates so they cannot drift apart: CheckOpenLong/Short (live entries),
the filtered-view sweep (a blocked side falls into the delete branch,
mirroring live), and the vote HUD's "-> TRADE" verdict. The verdict
re-derives at every label-cache rebuild, prints only on change, and is
computed even when the input is not Intelligent (marked informational).
NOT COMPILED - user compiles in MetaEditor.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
AnimateDread