1. Elliott Wave Scanner Indicator Description (MarkitTick)
Scans confirmed swing pivots for a complete 5-wave Impulse (1-2-3-4-5) or a complete 3-wave A-B-C Correction, validates the candidate against Elliott's structural rules and Fibonacci ratios, draws the wave count with a Fibonacci-based trade plan (entry, stop, target, Potential Reversal Zone), and reports pattern quality on a live scoring dashboard
1.1 Indicator Concept
Elliott Wave theory holds that price does not move randomly but in repeating 5-wave "impulse" swings (in the direction of the larger trend) followed by 3-wave "corrective" pullbacks (against it). This indicator does not try to predict a wave count in real time as it forms - instead it works backward from already-confirmed swing pivots. Every time enough confirmed pivots exist, it tests the most recent 6 pivots (3 highs + 3 lows) against the rules a genuine 5-wave impulse must obey, and separately tests the most recent 3 pivots (2 of one type + 1 of the other) against the ratio a genuine A-B-C correction must obey. Only when a candidate survives every rule and every enabled filter does it become a confirmed pattern - at which point the indicator uses the pattern's own geometry (the length of its completed legs) to project a Fibonacci-based entry, stop, and one or more targets, and shows a "Potential Reversal Zone" box marking where the next turn is statistically likely.
- Patterns are confirmed retrospectively, not predicted mid-formation: a pattern only appears once its final pivot (wave 5, or wave C) has already been confirmed by piv_right bars of price action - the indicator marks completed structure and proposes a forward trade idea from it, it does not forecast wave 3 while wave 2 is still forming.
- Two always-on rules, three toggleable rules: the Wave 2 retracement (38.2%-78.6% of Wave 1) and Wave 4 retracement (23.6%-50.0% of Wave 3) checks are applied unconditionally to every impulse candidate, while "Wave 3 not the shortest," "Wave 4 does not overlap Wave 1," and "Wave 3 extends Wave 1" can each be switched off independently in the Elliott Rules group.
- Only one pattern can be confirmed per bar, in a fixed priority order: Impulse Bull is tried first, then Impulse Bear, then Correction Bull, then Correction Bear - if an earlier candidate in that order qualifies, later candidates on the same bar are simply never drawn, even if they would also have passed their own rules.
- An Impulse always outranks a Correction covering the same bars: a newly confirmed 5-wave pattern actively deletes any previously drawn A-B-C pattern whose bar range overlaps it, while a Correction is blocked outright from being drawn if it would overlap an Impulse already on the chart - the relationship is one-directional.
1.2 Indicator Features
- Automatic Pivot Scanning: confirms swing highs/lows using a configurable left/right bar count, independently of any pattern logic.
- 5-Wave Impulse Detection (1-2-3-4-5): tests the most recent 6 pivots against the core Elliott impulse rules and Fibonacci retracement ranges, separately for bullish and bearish structures.
- 3-Wave Correction Detection (A-B-C): tests the most recent 3 pivots against a Fibonacci extension range for the C leg, separately for bullish and bearish structures, and labels the result "Flat" or "Exp Flat" depending on how far C traveled.
- Fibonacci Extension Lines (Impulse only): three dotted projection lines (1.0x / 1.618x / 2.618x of Wave 1's length) drawn forward from a confirmed impulse's end.
- Potential Reversal Zone (PRZ): a shaded box marking the Fibonacci price band where the pattern's next reversal is expected, drawn for both Impulse and Correction patterns.
- Trade Target Lines: dashed Entry / Stop / Target lines with price labels and a live Risk:Reward readout; Correction patterns additionally show a TP1 and TP3 line.
- Live Scoring Dashboard: a 22-row table reporting bias, pattern name/type, entry/stop/target, wave lengths and ratios, Risk:Reward quality, pivot count, and pattern count - all as visual "block bar" scores.
- Optional Bias Filters: Higher-Timeframe trend filter, Market-Structure midpoint filter, and Volume confirmation (Impulse only) - each independently toggleable.
- Pattern History Management: keeps only the most recent N patterns on the chart, automatically deleting the oldest pattern's drawings once the cap is exceeded.
- Alerts: separate alert conditions for a new Impulse and a new Correction, firing once per bar close with the pattern's entry/stop/target baked into the message.
1.3 How to Use the Indicator
- Read the dashboard's "R:R Quality" block bar as a shortcut for trade quality - it turns the raw Target/Stop distance ratio into a 0-10 score with color coding, so you can judge a setup at a glance without doing the Risk:Reward math yourself.
- Treat the PRZ box as a reaction zone, not a guaranteed reversal price - it is a Fibonacci-derived band, not a single line; price can pierce partway into it before actually turning.
- Remember an Impulse's Entry/Stop/Target is a continuation idea, not a reversal trade - the Target sits further in the same direction as Wave 5 (projected using 0.618 of Wave 3's length), and the Stop sits just beyond Wave 4 - so it is a trend-continuation setup taken right at the end of the marked 5-wave move, not a bet that price reverses once the count completes.
- Use a Correction pattern's three targets (TP1/TP2/TP3) to scale out - they are projected at 1.0x / 1.618x / 2.618x of the A-to-B leg's length beyond the C pivot, giving three progressively more ambitious continuation targets after the correction completes.
- Raise Pattern Cooldown if the same swing keeps re-triggering - as new bars form and the pivot window shifts by one, the same underlying structure can otherwise be re-confirmed and re-drawn repeatedly.
- Turn on the optional filters (HTF Bias, Market Structure, Volume) to cut down counter-trend or low-conviction signals - all three are off by default, so the base indicator will flag a valid pattern regardless of the larger trend or participation unless you opt in.
1.4 How the Indicator Works
Inputs & Their Roles
- piv_left / piv_right ("Pivot Left"/"Pivot Right", default: 5 / 5, range 2-50): bars required on each side of a high/low to confirm it as a pivot. A pivot only appears piv_right bars after it happened - raising either value makes pivots rarer and more significant, but delays confirmation further; e.g. at the default 5, a wave-5 pivot is only confirmed 5 bars after its actual high/low.
- max_piv ("Max Pivots", default: 500, range 10-1000): the maximum number of historical pivots kept in memory per type (high/low); has no effect on pattern logic itself, only on how far back the arrays can reach.
- cooldown ("Pattern Cooldown", default: 5, range 1-50): the minimum bar-index distance the new candidate's starting pivot must have from the last accepted pattern's starting pivot before a new pattern is accepted - without this, a slowly-shifting pivot window could re-confirm essentially the same wave count on consecutive bars.
- max_patterns ("Max Patterns on Chart", default: 3, range 1-20): how many completed patterns stay drawn simultaneously; once exceeded, the oldest pattern's lines/labels/boxes are deleted first.
- fib_tol ("Fib Tolerance (%)", default: 10.0%, range 1-30%, stored internally as a 0.01-0.30 decimal): a single tolerance applied to every Fibonacci ratio check in the script (Wave 2/4 retracement, Wave 5 ratio, Correction C extension). Raising it widens every ratio window - e.g. the Wave 2 retracement window of 38.2%-78.6% widens to roughly 34.4%-86.5% at the default 10% tolerance, so more (looser) candidates pass; lowering it makes the scanner far stricter and fewer patterns appear.
- min_size_atr ("Min Size (ATR x)", default: 0.8, range 0.1-10.0): the minimum length a wave leg must have, expressed as a multiple of the 14-period ATR, before it is considered large enough to matter. At the default 0.8, on an instrument with ATR = 10, any of Wave 1/3/5 (impulse) or the A/C legs (correction) shorter than 8 price units is rejected outright as noise.
- req_w3_ext ("W3 Must Extend W1", default: true): requires Wave 3's length to exceed Wave 1's length. Turning it off allows a valid impulse where Wave 3 is shorter than or equal to Wave 1 - a looser reading of the classic "Wave 3 is usually the longest" tendency.
- req_w4_no_ol ("W4 No Overlap W1", default: true): requires Wave 4's low (bullish) or high (bearish) to stay outside Wave 1's price territory - one of Elliott's three unbreakable rules in classic theory. Turning it off allows Wave 4 to retrace back into Wave 1's range, which is normally considered an invalidation of the impulse count.
- req_w3_not_short ("W3 Not Shortest", default: true): requires Wave 3's length to be at least as long as both Wave 1 and Wave 5 - the classic rule that Wave 3 (of 1, 3, 5) is never the shortest. Turning it off allows Wave 3 to be the shortest of the three impulse legs.
- req_w5_fib ("W5 Fib Ratio (61.8%-161.8% W1)", default: false): when enabled, additionally requires Wave 5's length to sit between 61.8% and 161.8% of Wave 1's length (within tolerance). Off by default, so Wave 5 can be any length as long as it clears the ATR size filter.
- req_b_fib ("ABC: B Retracement Filter", default: false): a toggle intended to constrain the B-leg retracement of a Correction pattern - see the implementation note below, as this filter has no real effect on which patterns pass.
- show_impulse / show_correction (default: true / true): independently enables detection and drawing of each pattern family; turning one off removes it from the priority order entirely, so the other family can only ever be confirmed.
- show_labels (default: true): toggles the "1"/"2"/"3"/"4" impulse-wave labels and the "A"/"B"/"C" correction labels (the pattern-name label in the center is drawn regardless of this setting).
- show_fib_lines (default: true): toggles the three dotted Fibonacci extension lines (Impulse patterns only).
- show_prz (default: true) / prz_transp ("PRZ Transparency", default: 88, range 50-99): toggles the Potential Reversal Zone box and sets how see-through it is - higher values are more transparent/subtle.
- show_targets (default: true) / stop_atr_buf ("SL ATR Buffer", default: 0.5): toggles the Entry/Stop/Target lines, and sets how far (in ATR multiples) the stop is placed beyond its anchor pivot - e.g. at ATR = 10 and the default 0.5, the stop sits 5 price units beyond Wave 4's low (impulse) or the C pivot (correction).
- min_rr ("Min R:R", default: 1.5, min 0.1): the minimum Reward:Risk ratio a candidate's projected trade must clear (Target vs. Entry/Stop distance for Impulse; TP2 vs. Entry/Stop distance for Correction) - candidates that fail this check are discarded entirely, even if every wave/Fibonacci rule passed.
- use_htf_bias (default: false) / htf_tf ("HTF TF", default: "240" = 4-hour): when enabled, only allows bullish patterns if the higher-timeframe close sits above its 50-period SMA (and only bearish patterns if it sits below) - both read one HTF bar back to avoid look-ahead.
- use_ms_filter (default: false) / ms_len ("MS Length", default: 20, range 5-100): when enabled, only allows bullish patterns if price closes above the midpoint of the last ms_len bars' high/low range (and only bearish patterns if below it).
- use_vol_confirm (default: false) / vol_len ("Vol SMA Len", default: 20): when enabled, requires the current bar's volume to exceed its vol_len-period average - checked only inside Impulse detection, never inside Correction detection.
- col_impulse / col_imp_bear / col_correction (defaults: cyan / red / orange): the wave color themes - note Correction patterns share one color for both bullish and bearish structures, unlike Impulse patterns which have distinct bull/bear colors.
- col_entry / col_stop / col_target / col_fib (defaults: yellow / pink / green / purple): fixed colors for the trade-target lines and Fibonacci extension lines, applied the same way regardless of pattern direction.
- alert_impulse / alert_correction (default: true / true): independently enable the "once per bar close" alert fired when a new pattern of that family is accepted.
Main Logic Flows
🎯 Flow 1: What You See
- A colored 5-leg zigzag (Impulse) or 2-leg zigzag (A-B-C Correction) connecting the relevant pivots, numbered/lettered wave labels, a pattern-name label, an optional shaded PRZ box, dashed Entry/Stop/Target lines with price + Risk:Reward labels, and a top-right dashboard table summarizing the latest pattern and live scan statistics.
🔍 Flow 2: Pivot Detection
- Every bar, a custom pivot-high/low check (f_pivotHigh / f_pivotLow) looks piv_left bars back and piv_right bars forward from a candidate bar; if every one of those surrounding bars is strictly less extreme, the candidate is confirmed as a pivot and inserted at the front of a running list (capped at max_piv entries per type, oldest dropped first).
📈 Flow 3: Impulse Pattern Validation
- Once at least 3 highs and 3 lows exist, the most recent 6 alternate pivots are taken as candidate Wave 0-1-2-3-4-5 points. The candidate must first pass a strict time-order and direction check, then: Wave 2 must retrace 38.2%-78.6% of Wave 1 (always on); Wave 4 must retrace 23.6%-50.0% of Wave 3 (always on); Wave 4 must not overlap Wave 1's price territory (toggle); Wave 3 must not be the shortest of Waves 1/3/5 (toggle); Wave 3 must exceed Wave 1's length (toggle); Wave 5 may optionally need to sit within 61.8%-161.8% of Wave 1 (toggle); and Waves 1, 3, 5 must each individually clear the ATR size filter. If Volume Confirmation is enabled, the current bar's volume must also exceed its SMA.
🔄 Flow 4: Correction (A-B-C) Pattern Validation
- Once at least 2 highs and 2 lows exist, the most recent 3 alternate pivots become candidate A-B-C points (e.g. for a bullish reading: A = older low, B = high in between, C = most recent low). Both the A-to-B leg and the B-to-C leg must clear the ATR size filter, C must extend 61.8%-161.8% of the A-to-B leg's length beyond B, and C must land beyond A in the continuation direction (at or below A for a bullish reading). The correction is then labeled "Flat" or "Exp Flat" depending on whether that C-extension ratio is above or below 1.0 (see the implementation note below regarding the B-leg filter and the unreachable "Zigzag" label).
🧭 Flow 5: Optional Bias Filters
- A candidate that passes its own wave rules is then checked, if enabled, against the Higher-Timeframe bias (HTF close vs. HTF 50-SMA, one bar back) and the Market-Structure midpoint filter (current close vs. the midpoint of the last ms_len bars' range) - bullish candidates need a bullish reading on every enabled filter, bearish candidates need a bearish reading on every enabled filter. Finally, the projected trade's Reward:Risk ratio is computed and must meet min_rr.
🏁 Flow 6: Priority Gate, Overlap Check & Cooldown
- Detection only runs past a warm-up point (bar_index beyond twice the pivot left+right window). Of the up-to-4 candidates computed this bar (Impulse Bull, Impulse Bear, Correction Bull, Correction Bear), only the first one in that fixed priority order that is enabled and valid is kept - later candidates on the same bar are discarded even if valid. A Correction candidate is additionally dropped if its bar range overlaps an Impulse pattern already drawn on the chart. The surviving candidate is then only actually accepted on a confirmed bar close, and only if no pattern has been accepted yet or the new candidate's starting pivot is more than cooldown bars away from the last accepted pattern's starting pivot.
✏️ Flow 7: Drawing the Pattern
- For an accepted Impulse: 5 connecting lines are drawn (Waves 1, 3, 5 in full wave color; Waves 2, 4 lighter/faded), labels "1"-"4" are placed at their respective pivots (Wave 5's own pivot is not separately labeled), a centered pattern-name label is added, and - if enabled - any existing Correction pattern whose bar range overlaps this new Impulse is deleted. If Fibonacci extension lines are enabled, three dotted lines (at 1.0x / 1.618x / 2.618x of Wave 1's length beyond Wave 4) are projected 50 bars forward with matching labels. For an accepted Correction: 2 connecting lines (A-to-B full color, B-to-C lighter) are drawn, labels "A"/"B"/"C" are placed, and a centered label shows the pattern name plus its Flat/Exp Flat classification in brackets.
🎯 Flow 8: PRZ Zone & Trade Target Lines
- If enabled, a shaded Potential Reversal Zone box is drawn spanning the pattern's own Fibonacci-projected reversal band, plus a "PRZ" label. If trade targets are enabled and valid, dashed Entry/Stop/Target lines are drawn extending 50 bars forward with price labels (the Target label also shows the live Risk:Reward ratio); for Correction patterns specifically, two further dashed lines (TP1 at 1.0x, TP3 at 2.618x of the A-to-B leg length beyond Entry) are added alongside the main Target line (which itself represents TP2 at 1.618x).
🗑️ Flow 9: Pattern History Management
- Every newly accepted pattern's drawings are pushed into a running history list; once that list exceeds max_patterns, the oldest pattern's lines, labels, and boxes are deleted from the chart entirely (not just hidden) to keep the drawing count bounded.
🔔 Flow 10: Alerts
- On the same confirmed bar a pattern is accepted, if the matching alert toggle (alert_impulse or alert_correction) is on, one alert fires (at most once per bar close) with the pattern's name, entry, stop, and target formatted into the message text.
📋 Flow 11: Dashboard
- Every bar, the dashboard table refreshes: bias (Bull/Bear), the latest pattern's name and type, its Entry/Stop/Target, a Risk:Reward "block bar" score (0-10, color-graded), Wave 1/3/5 lengths and their ratios to Wave 1, the correction type (if applicable), a total pivot-count "block bar" score, bars elapsed since the last pattern, the current ATR(14), the running Impulse+Correction pattern counters, and (if enabled) the HTF bias reading.
Implementation note - the "ABC: B Retracement Filter" has no real effect: the ratio
it checks, _b_wave_ret, is computed as
(wb - wa) / (wb - wa) - the same difference divided by itself - so
it is mathematically always exactly 1.0, regardless of the actual pivot prices. Since 1.0
comfortably sits inside the filter's accepted range under any realistic tolerance setting,
enabling req_b_fib never actually rejects a candidate. The same
self-referential ratio also drives the Flat/Exp Flat classification: because it is always
1.0, the "Zigzag" label that the code initializes as a default is never actually reached -
every confirmed Correction is labeled either "Flat" or "Exp Flat," decided purely by whether
the (separately, correctly computed) C-extension ratio is above or below 1.0.
Implementation note - a couple of smaller asymmetries: for Impulse patterns, only
Waves 1-4 receive a numbered label on the chart ("1" through "4") - Wave 5's own pivot point
is left unlabeled, even though it is the pivot the Entry line is anchored to. Separately,
Correction patterns use a single color (col_correction) for both
bullish and bearish structures, while Impulse patterns get two distinct colors
(col_impulse for bullish, col_imp_bear
for bearish) - so on the chart, direction is visually obvious for impulses but must be read
from the dashboard's Bias row or the "Bull"/"Bear" wording in the pattern name for
corrections.
Outputs & Usage Roles
- Wave count drawings (zigzag + labels): a visual record of exactly which pivots the indicator used to justify the pattern, letting you sanity-check the count against your own reading of the chart.
- Fibonacci extension lines (Impulse): forward-projected reference levels (1.0x/1.618x/2.618x of Wave 1) for gauging how far a continuation move could reasonably extend.
- PRZ box: a Fibonacci-derived price band flagging where the pattern's structure suggests the next reversal is likely to occur.
- Entry/Stop/Target lines (+ TP1/TP3 for Corrections): a ready-made, Fibonacci-based trade plan anchored to the just-confirmed pattern, with a live Risk:Reward readout.
- Dashboard: an at-a-glance quality and context summary - bias, pattern identity, trade levels, wave proportions, and how much history the scan has accumulated - without needing to inspect every drawing on the chart.
- Alerts: push notifications the moment a new Impulse or Correction is confirmed, carrying the same entry/stop/target the chart just drew.
// This work is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
// https://creativecommons.org/licenses/by-nc-sa/4.0/
// © MarkitTick
//@version=6
indicator("Elliott Wave Scanner [MarkitTick]", overlay=true, max_lines_count=500, max_labels_count=500, max_boxes_count=200, max_bars_back=1000)
// Inputs
string GRP_GEN = "General"
string GRP_PIV = "Pivot"
string GRP_EW = "Elliott Rules"
string GRP_COR = "Correction"
string GRP_FLT = "Filters"
string GRP_VIS = "Vis"
string GRP_DASH = "📊 Dash"
string GRP_ALT = "🔔 Alerts"
piv_left = input.int(5, "Pivot Left", minval=2, maxval=50, group=GRP_PIV)
piv_right = input.int(5, "Pivot Right", minval=2, maxval=50, group=GRP_PIV)
max_piv = input.int(500, "Max Pivots", minval=10, maxval=1000,group=GRP_PIV)
cooldown = input.int(5, "Pattern Cooldown", minval=1, maxval=50, group=GRP_GEN)
max_patterns = input.int(3, "Max Patterns on Chart",minval=1, maxval=20, group=GRP_GEN)
fib_tol = input.float(10.0,"Fib Tolerance (%)", minval=1.0,maxval=30.0,step=0.5,group=GRP_EW) / 100.0
min_size_atr = input.float(0.8, "Min Size (ATR×)", minval=0.1,maxval=10.0,step=0.1,group=GRP_EW)
req_w3_ext = input.bool(true, "W3 Must Extend W1", group=GRP_EW, tooltip="")
req_w4_no_ol = input.bool(true, "W4 No Overlap W1", group=GRP_EW, tooltip="")
req_w3_not_short= input.bool(true, "W3 Not Shortest", group=GRP_EW, tooltip="")
req_w5_fib = input.bool(false, "W5 Fib Ratio (61.8%–161.8% W1)",group=GRP_EW, tooltip="")
req_b_fib = input.bool(false, "ABC: B Retracement Filter", group=GRP_EW, tooltip="")
show_impulse = input.bool(true, "Show Impulse (12345)", group=GRP_VIS)
show_correction = input.bool(true, "Show Correction (ABC)", group=GRP_VIS)
show_labels = input.bool(true, "Show Wave Labels", group=GRP_VIS)
show_fib_lines = input.bool(true, "Show Fib Extension Lines", group=GRP_VIS)
show_prz = input.bool(true, "Show PRZ Zone", group=GRP_VIS)
prz_transp = input.int(88, "PRZ Transparency", minval=50,maxval=99, group=GRP_VIS)
show_targets = input.bool(true, "Show Trade Targets", group=GRP_VIS)
stop_atr_buf = input.float(0.5, "SL ATR Buffer", minval=0.0,step=0.1, group=GRP_VIS)
min_rr = input.float(1.5, "Min R:R", minval=0.1,step=0.1, group=GRP_FLT)
use_htf_bias = input.bool(false, "HTF Bias Filter", group=GRP_FLT, tooltip="")
htf_tf = input.timeframe("240","HTF TF", group=GRP_FLT)
use_ms_filter = input.bool(false, "Market Structure Filter", group=GRP_FLT, tooltip="")
ms_len = input.int(20, "MS Length", minval=5,maxval=100, group=GRP_FLT)
use_vol_confirm = input.bool(false, "Volume Confirmation W3", group=GRP_FLT, tooltip="")
vol_len = input.int(20, "Vol SMA Len", group=GRP_FLT)
col_impulse = input.color(color.rgb(0, 200, 255, 0), "Impulse Bull Color", group=GRP_VIS)
col_imp_bear = input.color(color.rgb(255, 80, 80, 0), "Impulse Bear Color", group=GRP_VIS)
col_correction = input.color(color.rgb(255, 160, 0, 0), "Correction Color", group=GRP_VIS)
col_entry = input.color(color.rgb(255, 220, 0, 0), "Entry Color", group=GRP_VIS)
col_stop = input.color(color.rgb(255, 50, 130, 0), "Stop Color", group=GRP_VIS)
col_target = input.color(color.rgb(0, 255, 160, 0), "Target Color", group=GRP_VIS)
col_fib = input.color(color.rgb(180, 100, 255, 0), "Fib Line Color", group=GRP_VIS)
color dash_bg_color = input.color(color.rgb(10, 10, 18, 5), "BG", group=GRP_DASH)
color dash_txt_color = input.color(color.rgb(220, 220, 235, 0), "Text", group=GRP_DASH)
bool alert_impulse = input.bool(true, "Alert: Impulse Detected", group=GRP_ALT)
bool alert_correction = input.bool(true, "Alert: Correction Detected", group=GRP_ALT)
// Types
type Pivot
float p
int i
type WaveResult
bool detected
string name
bool isBullish
float w1_lo
float w1_hi
int w1_lo_i
int w1_hi_i
float w2_p
int w2_i
float w3_p
int w3_i
float w4_p
int w4_i
float w5_p
int w5_i
float wa_p
int wa_i
float wb_p
int wb_i
float wc_p
int wc_i
float entry_p
float stop_p
float target_p
float tp2_p
float tp3_p
float prz_lo
float prz_hi
int start_i
int end_i
float w1_len
float w3_len
float w5_len
int pattern_type
string corr_type
type WaveDrawings
line[] lns
label[] lbls
box[] bxs
int pattern_type
int start_i
int end_i
// State
var WaveResult EMPTY_W = WaveResult.new(false, "", false, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, na, 0, "")
var array<Pivot> pivotHighs = array.new<Pivot>()
var array<Pivot> pivotLows = array.new<Pivot>()
var array<WaveDrawings> pat_history = array.new<WaveDrawings>()
var string dash_name = "—"
var bool dash_bull = false
var int dash_last_bar = 0
var float dash_entry = na
var float dash_stop = na
var float dash_target = na
var float dash_w1_len = na
var float dash_w3_len = na
var float dash_w5_len = na
var float dash_w3_ratio = na
var float dash_w5_ratio = na
var float dash_fib_a = na
var float dash_fib_b = na
var float dash_fib_c = na
var int dash_pattern_type= 0
var string dash_corr_type = "—"
var float dash_tp2 = na
var float dash_tp3 = na
var int frozen_idx = na
var int impulse_count = 0
var int correction_count = 0
// Calculations
string clean_ticker = syminfo.prefix + ":" + syminfo.ticker
float atr_val = ta.atr(14)
float vol_sma = use_vol_confirm ? ta.sma(volume, math.max(1, math.min(vol_len, bar_index + 1))) : na
[htf_close, htf_ema] = if use_htf_bias
request.security(syminfo.tickerid, htf_tf, [close[1], ta.sma(close, 50)[1]], lookahead=barmerge.lookahead_on)
else
[na, na]
bool htf_bull = use_htf_bias ? htf_close > htf_ema : true
bool htf_bear = use_htf_bias ? htf_close < htf_ema : true
float ms_hi = use_ms_filter ? ta.highest(high, ms_len) : na
float ms_lo = use_ms_filter ? ta.lowest(low, ms_len) : na
float ms_mid = (ms_hi + ms_lo) * 0.5
bool ms_bull = use_ms_filter and close > ms_mid
bool ms_bear = use_ms_filter and close < ms_mid
int dash_bars_since = bar_index - dash_last_bar
float dash_rr_raw = not na(dash_entry) and not na(dash_stop) and not na(dash_target) and math.abs(dash_entry - dash_stop) > 1e-10 ? math.abs(dash_target - dash_entry) / math.abs(dash_entry - dash_stop) : 0.0
f_blockBar(float score) =>
int n = math.min(math.max(math.round(score), 0), 10)
string s = ""
for i = 1 to 10
s += i <= n ? "█" : "░"
s + " " + str.tostring(math.round(score / 10.0 * 100.0), "##0") + "%"
f_blockColor(float score) =>
score >= 7.0 ? color.rgb(0, 255, 120) : score >= 4.0 ? color.rgb(80, 255, 0) : score >= 1.5 ? color.rgb(200, 200, 0) : color.rgb(110, 110, 110)
f_pivotHigh(series float src, simple int llen, simple int rlen) =>
float candidate = src[rlen]
bool isValid = true
for i = 1 to llen + rlen
if i != rlen and src[i] >= candidate
isValid := false
isValid ? candidate : na
f_pivotLow(series float src, simple int llen, simple int rlen) =>
float candidate = src[rlen]
bool isValid = true
for i = 1 to llen + rlen
if i != rlen and src[i] <= candidate
isValid := false
isValid ? candidate : na
f_inRange(float ratio, float lo, float hi, float tol) =>
ratio >= lo * (1.0 - tol) and ratio <= hi * (1.0 + tol)
f_fibRatio(float leg1, float leg2) =>
math.abs(leg1) > 1e-10 ? math.abs(leg2) / math.abs(leg1) : na
f_isValidSize(float h) =>
h > atr_val * min_size_atr
f_sl(float anchor, bool isBull) =>
isBull ? anchor - atr_val * stop_atr_buf : anchor + atr_val * stop_atr_buf
f_validate_rr(float entry_p, float stop_p, float target_p) =>
float risk = math.abs(entry_p - stop_p)
float reward = math.abs(target_p - entry_p)
risk > 1e-10 and reward / risk >= min_rr
f_ms_bias(bool isBull) => not use_ms_filter or (isBull ? ms_bull : ms_bear)
f_htf_bias(bool isBull) => not use_htf_bias or (isBull ? htf_bull : htf_bear)
f_manage_drawings(int max_items) =>
while array.size(pat_history) > max_items
WaveDrawings old = array.shift(pat_history)
int _ls = array.size(old.lns)
if _ls > 0
for i = 0 to _ls - 1
line _l = array.get(old.lns, i)
if not na(_l)
line.delete(_l)
int _lb = array.size(old.lbls)
if _lb > 0
for i = 0 to _lb - 1
label _l = array.get(old.lbls, i)
if not na(_l)
label.delete(_l)
int _bx = array.size(old.bxs)
if _bx > 0
for i = 0 to _bx - 1
box _b = array.get(old.bxs, i)
if not na(_b)
box.delete(_b)
f_deleteDrawing(WaveDrawings d) =>
int _ls = array.size(d.lns)
if _ls > 0
for i = 0 to _ls - 1
line _l = array.get(d.lns, i)
if not na(_l)
line.delete(_l)
int _lb = array.size(d.lbls)
if _lb > 0
for i = 0 to _lb - 1
label _l = array.get(d.lbls, i)
if not na(_l)
label.delete(_l)
int _bx = array.size(d.bxs)
if _bx > 0
for i = 0 to _bx - 1
box _b = array.get(d.bxs, i)
if not na(_b)
box.delete(_b)
f_purgeOverlappingCorrections(int newStart, int newEnd) =>
int _n = array.size(pat_history)
if _n > 0
array<int> _toRemove = array.new<int>()
for i = 0 to _n - 1
WaveDrawings _d = array.get(pat_history, i)
bool _isCorrection = _d.pattern_type == 2
bool _overlap = _d.start_i <= newEnd and _d.end_i >= newStart
if _isCorrection and _overlap
array.push(_toRemove, i)
int _rn = array.size(_toRemove)
if _rn > 0
for j = _rn - 1 to 0
int _ri = array.get(_toRemove, j)
WaveDrawings _dd = array.get(pat_history, _ri)
f_deleteDrawing(_dd)
array.remove(pat_history, _ri)
f_overlapsExistingImpulse(int newStart, int newEnd) =>
bool _found = false
int _n = array.size(pat_history)
if _n > 0
for i = 0 to _n - 1
WaveDrawings _d = array.get(pat_history, i)
bool _isImpulse = _d.pattern_type == 1
bool _overlap = _d.start_i <= newEnd and _d.end_i >= newStart
if _isImpulse and _overlap
_found := true
_found
// Signals
detectImpulseBull(array<Pivot> highs, array<Pivot> lows) =>
res = EMPTY_W
bool _has = highs.size() >= 3 and lows.size() >= 3
if not _has
res
else
_p0 = lows.get(2)
_p1 = highs.get(2)
_p2 = lows.get(1)
_p3 = highs.get(1)
_p4 = lows.get(0)
_p5 = highs.get(0)
bool _seq = _p0.i < _p1.i and _p1.i < _p2.i and _p2.i < _p3.i and _p3.i < _p4.i and _p4.i < _p5.i
bool _dir = _p1.p > _p0.p and _p2.p > _p0.p and _p3.p > _p1.p
bool _w2 = _p2.p > _p0.p
float _w1_len = _p1.p - _p0.p
float _w2_ret = _p1.p > _p0.p ? (_p1.p - _p2.p) / _w1_len : na
float _w3_len = _p3.p - _p2.p
float _w4_ret = _p3.p > _p2.p ? (_p3.p - _p4.p) / _w3_len : na
float _w5_len = _p5.p - _p4.p
bool _w4_no_ol = not req_w4_no_ol or (_p4.p > _p1.p)
bool _w3_not_s = not req_w3_not_short or (_w3_len >= _w1_len and _w3_len >= _w5_len)
bool _w3_ext = not req_w3_ext or (_w3_len > _w1_len)
float _w5_rat = _w1_len > 1e-10 ? _w5_len / _w1_len : na
bool _w5_fib = not req_w5_fib or (not na(_w5_rat) and f_inRange(_w5_rat, 0.618, 1.618, fib_tol))
bool _w2_fib = not na(_w2_ret) and f_inRange(_w2_ret, 0.382, 0.786, fib_tol)
bool _w4_fib = not na(_w4_ret) and f_inRange(_w4_ret, 0.236, 0.500, fib_tol)
bool _w5_valid = f_isValidSize(_w5_len) and f_isValidSize(_w1_len) and f_isValidSize(_w3_len)
bool _vol_ok = not use_vol_confirm or (not na(vol_sma) and volume > vol_sma)
if _seq and _dir and _w2 and _w2_fib and _w4_fib and _w4_no_ol and _w3_not_s and _w3_ext and _w5_fib and _w5_valid and _vol_ok
if f_ms_bias(true) and f_htf_bias(true)
float _w5_proj_161 = _p4.p + _w1_len * 1.618
float _w5_proj_100 = _p4.p + _w1_len * 1.000
float _prz_lo = math.min(_w5_proj_100, _p5.p)
float _prz_hi = math.max(_w5_proj_161, _p5.p)
float _entry_p = _p5.p
float _stop_p = f_sl(_p4.p, true)
float _target_p = _p5.p + _w3_len * 0.618
bool _rr_ok = f_validate_rr(_entry_p, _stop_p, _target_p)
if _rr_ok
res := WaveResult.new(true, "Impulse Bull", true, _p0.p, _p1.p, _p0.i, _p1.i, _p2.p, _p2.i, _p3.p, _p3.i, _p4.p, _p4.i, _p5.p, _p5.i, na, na, na, na, na, na, _entry_p, _stop_p, _target_p, na, na, _prz_lo, _prz_hi, _p0.i, _p5.i, _w1_len, _w3_len, _w5_len, 1, "")
res
detectImpulseBear(array<Pivot> highs, array<Pivot> lows) =>
res = EMPTY_W
bool _has = highs.size() >= 3 and lows.size() >= 3
if not _has
res
else
_p0 = highs.get(2)
_p1 = lows.get(2)
_p2 = highs.get(1)
_p3 = lows.get(1)
_p4 = highs.get(0)
_p5 = lows.get(0)
bool _seq = _p0.i < _p1.i and _p1.i < _p2.i and _p2.i < _p3.i and _p3.i < _p4.i and _p4.i < _p5.i
bool _dir = _p1.p < _p0.p and _p2.p < _p0.p and _p3.p < _p1.p
bool _w2 = _p2.p < _p0.p
float _w1_len = _p0.p - _p1.p
float _w2_ret = _p0.p > _p1.p ? (_p2.p - _p1.p) / _w1_len : na
float _w3_len = _p2.p - _p3.p
float _w4_ret = _p2.p > _p3.p ? (_p4.p - _p3.p) / _w3_len : na
float _w5_len = _p4.p - _p5.p
bool _w4_no_ol = not req_w4_no_ol or (_p4.p < _p1.p)
bool _w3_not_s = not req_w3_not_short or (_w3_len >= _w1_len and _w3_len >= _w5_len)
bool _w3_ext = not req_w3_ext or (_w3_len > _w1_len)
float _w5_rat = _w1_len > 1e-10 ? _w5_len / _w1_len : na
bool _w5_fib = not req_w5_fib or (not na(_w5_rat) and f_inRange(_w5_rat, 0.618, 1.618, fib_tol))
bool _w2_fib = not na(_w2_ret) and f_inRange(_w2_ret, 0.382, 0.786, fib_tol)
bool _w4_fib = not na(_w4_ret) and f_inRange(_w4_ret, 0.236, 0.500, fib_tol)
bool _w5_valid = f_isValidSize(_w5_len) and f_isValidSize(_w1_len) and f_isValidSize(_w3_len)
bool _vol_ok = not use_vol_confirm or (not na(vol_sma) and volume > vol_sma)
if _seq and _dir and _w2 and _w2_fib and _w4_fib and _w4_no_ol and _w3_not_s and _w3_ext and _w5_fib and _w5_valid and _vol_ok
if f_ms_bias(false) and f_htf_bias(false)
float _w5_proj_161 = _p4.p - _w1_len * 1.618
float _w5_proj_100 = _p4.p - _w1_len * 1.000
float _prz_lo = math.min(_w5_proj_161, _p5.p)
float _prz_hi = math.max(_w5_proj_100, _p5.p)
float _entry_p = _p5.p
float _stop_p = f_sl(_p4.p, false)
float _target_p = _p5.p - _w3_len * 0.618
bool _rr_ok = f_validate_rr(_entry_p, _stop_p, _target_p)
if _rr_ok
res := WaveResult.new(true, "Impulse Bear", false, _p1.p, _p0.p, _p1.i, _p0.i, _p2.p, _p2.i, _p3.p, _p3.i, _p4.p, _p4.i, _p5.p, _p5.i, na, na, na, na, na, na, _entry_p, _stop_p, _target_p, na, na, _prz_lo, _prz_hi, _p0.i, _p5.i, _w1_len, _w3_len, _w5_len, 1, "")
res
detectCorrectionBull(array<Pivot> highs, array<Pivot> lows) =>
res = EMPTY_W
bool _has = highs.size() >= 2 and lows.size() >= 2
if not _has
res
else
_wa = lows.get(1)
_wb = highs.get(1)
_wc = lows.get(0)
bool _seq = _wa.i < _wb.i and _wb.i < _wc.i
float _a_len = _wb.p - _wa.p
float _b_ret = _a_len > 1e-10 ? (_wb.p - _wc.p) / _a_len : na
float _c_drop = _wb.p - _wc.p
float _c_ext = _a_len > 1e-10 ? _c_drop / _a_len : na
float _b_ret_ab = _a_len > 1e-10 ? (_wb.p - _wa.p - (_wb.p - _wc.p)) / _a_len : na
float _b_from_a = _a_len > 1e-10 ? (_wb.p - _wc.p) / _a_len : na
bool _abc_ok = f_isValidSize(_a_len) and f_isValidSize(_c_drop)
bool _c_ok = not na(_c_ext) and f_inRange(_c_ext, 0.618, 1.618, fib_tol)
bool _c_below_a = _wc.p <= _wa.p
float _b_wave_ret = _a_len > 1e-10 ? (_wb.p - _wa.p) / _a_len : na
bool _b_ok = not req_b_fib or (not na(_b_wave_ret) and _b_wave_ret >= 0.382 * (1.0 - fib_tol) and _b_wave_ret <= 1.382 * (1.0 + fib_tol))
string _ctype = "Zigzag"
if not na(_b_wave_ret) and not na(_c_ext)
if _b_wave_ret >= 0.9 * (1.0 - fib_tol) and _b_wave_ret <= 1.05 * (1.0 + fib_tol)
_ctype := _c_ext > 1.0 ? "Exp Flat" : "Flat"
else
_ctype := "Zigzag"
if _seq and _abc_ok and _c_ok and _c_below_a and _b_ok
if f_ms_bias(true) and f_htf_bias(true)
float _prz_lo = _wb.p - _a_len * 1.618 * (1.0 + fib_tol)
float _prz_hi = _wb.p - _a_len * 0.618 * (1.0 - fib_tol)
float _entry_p = _wc.p
float _stop_p = f_sl(_wc.p, true)
float _tp1 = _wc.p + _a_len * 1.000
float _tp2 = _wc.p + _a_len * 1.618
float _tp3 = _wc.p + _a_len * 2.618
bool _rr_ok = f_validate_rr(_entry_p, _stop_p, _tp2)
if _rr_ok
res := WaveResult.new(true, "ABC Bull", true, na, na, na, na, na, na, na, na, na, na, na, na, _wa.p, _wa.i, _wb.p, _wb.i, _wc.p, _wc.i, _entry_p, _stop_p, _tp2, _tp2, _tp3, _prz_lo, _prz_hi, _wa.i, _wc.i, _a_len, na, na, 2, _ctype)
res
detectCorrectionBear(array<Pivot> highs, array<Pivot> lows) =>
res = EMPTY_W
bool _has = highs.size() >= 2 and lows.size() >= 2
if not _has
res
else
_wa = highs.get(1)
_wb = lows.get(1)
_wc = highs.get(0)
bool _seq = _wa.i < _wb.i and _wb.i < _wc.i
float _a_len = _wa.p - _wb.p
float _c_rise = _wc.p - _wb.p
float _c_ext = _a_len > 1e-10 ? _c_rise / _a_len : na
bool _abc_ok = f_isValidSize(_a_len) and f_isValidSize(_c_rise)
bool _c_ok = not na(_c_ext) and f_inRange(_c_ext, 0.618, 1.618, fib_tol)
bool _c_above_a = _wc.p >= _wa.p
float _b_wave_ret = _a_len > 1e-10 ? (_wa.p - _wb.p) / _a_len : na
bool _b_ok = not req_b_fib or (not na(_b_wave_ret) and _b_wave_ret >= 0.382 * (1.0 - fib_tol) and _b_wave_ret <= 1.382 * (1.0 + fib_tol))
string _ctype = "Zigzag"
if not na(_b_wave_ret) and not na(_c_ext)
if _b_wave_ret >= 0.9 * (1.0 - fib_tol) and _b_wave_ret <= 1.05 * (1.0 + fib_tol)
_ctype := _c_ext > 1.0 ? "Exp Flat" : "Flat"
else
_ctype := "Zigzag"
if _seq and _abc_ok and _c_ok and _c_above_a and _b_ok
if f_ms_bias(false) and f_htf_bias(false)
float _prz_lo = _wb.p + _a_len * 0.618 * (1.0 - fib_tol)
float _prz_hi = _wb.p + _a_len * 1.618 * (1.0 + fib_tol)
float _entry_p = _wc.p
float _stop_p = f_sl(_wc.p, false)
float _tp1 = _wc.p - _a_len * 1.000
float _tp2 = _wc.p - _a_len * 1.618
float _tp3 = _wc.p - _a_len * 2.618
bool _rr_ok = f_validate_rr(_entry_p, _stop_p, _tp2)
if _rr_ok
res := WaveResult.new(true, "ABC Bear", false, na, na, na, na, na, na, na, na, na, na, na, na, _wa.p, _wa.i, _wb.p, _wb.i, _wc.p, _wc.i, _entry_p, _stop_p, _tp2, _tp2, _tp3, _prz_lo, _prz_hi, _wa.i, _wc.i, _a_len, na, na, 2, _ctype)
res
float ph = f_pivotHigh(high, piv_left, piv_right)
float pl = f_pivotLow(low, piv_left, piv_right)
if not na(ph)
pivotHighs.unshift(Pivot.new(ph, bar_index - piv_right))
if pivotHighs.size() > max_piv
pivotHighs.pop()
if not na(pl)
pivotLows.unshift(Pivot.new(pl, bar_index - piv_right))
if pivotLows.size() > max_piv
pivotLows.pop()
bool has_imp = pivotHighs.size() >= 3 and pivotLows.size() >= 3
bool has_cor = pivotHighs.size() >= 2 and pivotLows.size() >= 2
WaveResult _r_imp_bull = has_imp ? detectImpulseBull(pivotHighs, pivotLows) : EMPTY_W
WaveResult _r_imp_bear = has_imp ? detectImpulseBear(pivotHighs, pivotLows) : EMPTY_W
WaveResult _r_cor_bull = has_cor ? detectCorrectionBull(pivotHighs, pivotLows) : EMPTY_W
WaveResult _r_cor_bear = has_cor ? detectCorrectionBear(pivotHighs, pivotLows) : EMPTY_W
WaveResult res_w = EMPTY_W
bool _gate = bar_index > (piv_left + piv_right) * 2
if _gate
if show_impulse and not res_w.detected and _r_imp_bull.detected
res_w := _r_imp_bull
if show_impulse and not res_w.detected and _r_imp_bear.detected
res_w := _r_imp_bear
bool _cor_bull_clear = show_correction and not res_w.detected and _r_cor_bull.detected and not f_overlapsExistingImpulse(_r_cor_bull.start_i, _r_cor_bull.end_i)
bool _cor_bear_clear = show_correction and not res_w.detected and _r_cor_bear.detected and not f_overlapsExistingImpulse(_r_cor_bear.start_i, _r_cor_bear.end_i)
if _cor_bull_clear
res_w := _r_cor_bull
if not res_w.detected and _cor_bear_clear
res_w := _r_cor_bear
bool _is_new = res_w.detected and barstate.isconfirmed and (na(frozen_idx) or math.abs(res_w.start_i - frozen_idx) > cooldown)
if _is_new
WaveDrawings cur = WaveDrawings.new(array.new_line(), array.new_label(), array.new_box(), res_w.pattern_type, res_w.start_i, res_w.end_i)
frozen_idx := res_w.start_i
dash_name := res_w.name
dash_bull := res_w.isBullish
dash_last_bar := bar_index
dash_entry := res_w.entry_p
dash_stop := res_w.stop_p
dash_target := res_w.target_p
dash_w1_len := res_w.w1_len
dash_w3_len := res_w.w3_len
dash_w5_len := res_w.w5_len
dash_pattern_type:= res_w.pattern_type
dash_corr_type := res_w.corr_type
dash_tp2 := res_w.tp2_p
dash_tp3 := res_w.tp3_p
dash_w5_ratio := not na(res_w.w1_len) and res_w.w1_len > 1e-10 and not na(res_w.w5_len) ? res_w.w5_len / res_w.w1_len : na
if res_w.pattern_type == 1
impulse_count := impulse_count + 1
else
correction_count := correction_count + 1
color _col = res_w.pattern_type == 1 ? (res_w.isBullish ? col_impulse : col_imp_bear) : col_correction
bool _is_impulse = res_w.pattern_type == 1
if _is_impulse
f_purgeOverlappingCorrections(res_w.start_i, res_w.end_i)
if _is_impulse and not na(res_w.w1_lo_i) and not na(res_w.w5_i)
int _x0 = res_w.isBullish ? res_w.w1_lo_i : res_w.w1_hi_i
float _y0 = res_w.isBullish ? res_w.w1_lo : res_w.w1_hi
int _x1 = res_w.isBullish ? res_w.w1_hi_i : res_w.w1_lo_i
float _y1 = res_w.isBullish ? res_w.w1_hi : res_w.w1_lo
int _x2 = res_w.w2_i
float _y2 = res_w.w2_p
int _x3 = res_w.w3_i
float _y3 = res_w.w3_p
int _x4 = res_w.w4_i
float _y4 = res_w.w4_p
int _x5 = res_w.w5_i
float _y5 = res_w.w5_p
line _l01 = line.new(_x0, _y0, _x1, _y1, color=_col, width=3)
line _l12 = line.new(_x1, _y1, _x2, _y2, color=color.new(_col, 40), width=3)
line _l23 = line.new(_x2, _y2, _x3, _y3, color=_col, width=3)
line _l34 = line.new(_x3, _y3, _x4, _y4, color=color.new(_col, 40), width=3)
line _l45 = line.new(_x4, _y4, _x5, _y5, color=_col, width=3)
array.push(cur.lns, _l01)
array.push(cur.lns, _l12)
array.push(cur.lns, _l23)
array.push(cur.lns, _l34)
array.push(cur.lns, _l45)
if show_fib_lines and not na(res_w.w1_len) and res_w.w1_len > 1e-10
float _fib_161 = res_w.isBullish ? _y4 + res_w.w1_len * 1.618 : _y4 - res_w.w1_len * 1.618
float _fib_100 = res_w.isBullish ? _y4 + res_w.w1_len * 1.000 : _y4 - res_w.w1_len * 1.000
float _fib_262 = res_w.isBullish ? _y4 + res_w.w1_len * 2.618 : _y4 - res_w.w1_len * 2.618
int _fx1 = res_w.end_i
int _fx2 = res_w.end_i + 50
int _flx = _fx2 + 1
line _lf1 = line.new(_fx1, _fib_100, _fx2, _fib_100, color=color.new(col_fib, 60), style=line.style_dotted, width=2)
line _lf2 = line.new(_fx1, _fib_161, _fx2, _fib_161, color=color.new(col_fib, 40), style=line.style_dotted, width=2)
line _lf3 = line.new(_fx1, _fib_262, _fx2, _fib_262, color=color.new(col_fib, 70), style=line.style_dotted, width=2)
label _lf1l = label.new(_flx, _fib_100, "W5 1.0×", color=color.new(col_fib, 80), textcolor=col_fib, style=label.style_label_left, size=size.small)
label _lf2l = label.new(_flx, _fib_161, "W5 1.618×", color=color.new(col_fib, 80), textcolor=col_fib, style=label.style_label_left, size=size.small)
label _lf3l = label.new(_flx, _fib_262, "W5 2.618×", color=color.new(col_fib, 80), textcolor=col_fib, style=label.style_label_left, size=size.small)
array.push(cur.lns, _lf1)
array.push(cur.lns, _lf2)
array.push(cur.lns, _lf3)
array.push(cur.lbls, _lf1l)
array.push(cur.lbls, _lf2l)
array.push(cur.lbls, _lf3l)
if show_labels
label _lb0 = label.new(_x1, _y1, "1", color=color.new(_col, 100), textcolor=_col, style=res_w.isBullish ? label.style_label_down : label.style_label_up, size=size.small)
label _lb1 = label.new(_x2, _y2, "2", color=color.new(_col, 100), textcolor=_col, style=res_w.isBullish ? label.style_label_up : label.style_label_down, size=size.small)
label _lb2 = label.new(_x3, _y3, "3", color=color.new(_col, 100), textcolor=_col, style=res_w.isBullish ? label.style_label_down : label.style_label_up, size=size.small)
label _lb3 = label.new(_x4, _y4, "4", color=color.new(_col, 100), textcolor=_col, style=res_w.isBullish ? label.style_label_up : label.style_label_down, size=size.small)
array.push(cur.lbls, _lb0)
array.push(cur.lbls, _lb1)
array.push(cur.lbls, _lb2)
array.push(cur.lbls, _lb3)
int _cx = (res_w.start_i + res_w.end_i) / 2
float _cy = res_w.isBullish ? (nz(res_w.w3_p) + nz(res_w.w1_lo)) * 0.5 : (nz(res_w.w3_p) + nz(res_w.w1_hi)) * 0.5
label _lbl_name = label.new(_cx, _cy, res_w.name, color=color.rgb(0, 0, 0, 0), textcolor=_col, style=label.style_label_center, size=size.small)
array.push(cur.lbls, _lbl_name)
if not _is_impulse and not na(res_w.wa_i) and not na(res_w.wc_i)
line _la = line.new(res_w.wa_i, res_w.wa_p, res_w.wb_i, res_w.wb_p, color=_col, width=2)
line _lb2 = line.new(res_w.wb_i, res_w.wb_p, res_w.wc_i, res_w.wc_p, color=color.new(_col, 30), width=1)
array.push(cur.lns, _la)
array.push(cur.lns, _lb2)
if show_labels
label _lwa = label.new(res_w.wa_i, res_w.wa_p, "A", color=color.new(_col, 20), textcolor=color.white, style=res_w.isBullish ? label.style_label_down : label.style_label_up, size=size.small)
label _lwb = label.new(res_w.wb_i, res_w.wb_p, "B", color=color.new(_col, 20), textcolor=color.white, style=res_w.isBullish ? label.style_label_up : label.style_label_down, size=size.small)
label _lwc = label.new(res_w.wc_i, res_w.wc_p, "C", color=color.new(_col, 20), textcolor=color.white, style=res_w.isBullish ? label.style_label_down : label.style_label_up, size=size.small)
array.push(cur.lbls, _lwa)
array.push(cur.lbls, _lwb)
array.push(cur.lbls, _lwc)
int _cx2 = (res_w.wa_i + res_w.wc_i) / 2
float _cy2 = (res_w.wa_p + res_w.wc_p) * 0.5
string _abc_lbl_txt = res_w.name + (res_w.corr_type != "" ? " [" + res_w.corr_type + "]" : "")
label _lbl_n2 = label.new(_cx2, _cy2, _abc_lbl_txt, color=color.rgb(0, 0, 0, 0), textcolor=color.white, style=label.style_label_center, size=size.small)
array.push(cur.lbls, _lbl_n2)
if show_prz and not na(res_w.prz_lo) and not na(res_w.prz_hi)
int _prz_end = res_w.end_i
int _prz_x1 = _prz_end - 3
int _prz_x2 = _prz_end + 40
box _prz = box.new(_prz_x1, res_w.prz_hi, _prz_x2, res_w.prz_lo, bgcolor=color.new(_col, prz_transp), border_color=color.new(_col, 40), border_width=1)
array.push(cur.bxs, _prz)
label _prz_lbl = label.new(_prz_x2, (res_w.prz_hi + res_w.prz_lo) * 0.5, "PRZ", style=label.style_label_left, color=color.new(_col, 50), textcolor=_col, size=size.small)
array.push(cur.lbls, _prz_lbl)
if show_targets and not na(res_w.entry_p) and not na(res_w.stop_p) and not na(res_w.target_p)
int _x1 = res_w.end_i
int _x2 = res_w.end_i + 50
int _lx = _x2 + 1
float _ep = res_w.entry_p
float risk_v = math.abs(_ep - res_w.stop_p)
float reward_v = math.abs(res_w.target_p - _ep)
float rr_v = risk_v > 1e-10 ? reward_v / risk_v : 0.0
string _ep_s = str.tostring(_ep, format.mintick)
string _sp_s = str.tostring(res_w.stop_p, format.mintick)
string _tp_s = str.tostring(res_w.target_p, format.mintick)
string _rr_s = str.tostring(rr_v, "#.##")
line ln_e = line.new(_x1, _ep, _x2, _ep, color=col_entry, style=line.style_dashed, width=1)
line ln_s = line.new(_x1, res_w.stop_p, _x2, res_w.stop_p, color=col_stop, style=line.style_dashed, width=1)
line ln_t = line.new(_x1, res_w.target_p,_x2, res_w.target_p,color=col_target, style=line.style_dashed, width=1)
label lb_e = label.new(_lx, _ep, "▶ Entry " + _ep_s, style=label.style_label_left, color=color.new(col_entry, 70), textcolor=col_entry, size=size.small)
label lb_s = label.new(_lx, res_w.stop_p, "▶ Stop " + _sp_s, style=label.style_label_left, color=color.new(col_stop, 70), textcolor=col_stop, size=size.small)
label lb_t = label.new(_lx, res_w.target_p, "▶ Target " + _tp_s + " R:R " + _rr_s, style=label.style_label_left, color=color.new(col_target, 70), textcolor=col_target, size=size.small)
array.push(cur.lns, ln_e)
array.push(cur.lns, ln_s)
array.push(cur.lns, ln_t)
array.push(cur.lbls, lb_e)
array.push(cur.lbls, lb_s)
array.push(cur.lbls, lb_t)
if not _is_impulse and not na(res_w.tp2_p) and not na(res_w.tp3_p)
float _a_base = res_w.w1_len
float _tp1_p = res_w.isBullish ? res_w.entry_p + _a_base * 1.000 : res_w.entry_p - _a_base * 1.000
line ln_t1 = line.new(_x1, _tp1_p, _x2, _tp1_p, color=color.new(col_target, 60), style=line.style_dashed, width=1)
line ln_t3 = line.new(_x1, res_w.tp3_p, _x2, res_w.tp3_p, color=color.new(col_target, 20), style=line.style_dashed, width=1)
label lb_t1 = label.new(_lx, _tp1_p, "▶ TP1 A×1.0 " + str.tostring(_tp1_p, format.mintick), style=label.style_label_left, color=color.new(col_target, 70), textcolor=color.new(col_target, 30), size=size.small)
label lb_t3 = label.new(_lx, res_w.tp3_p, "▶ TP3 A×2.618 " + str.tostring(res_w.tp3_p, format.mintick), style=label.style_label_left, color=color.new(col_target, 70), textcolor=color.new(col_target, 10), size=size.small)
array.push(cur.lns, ln_t1)
array.push(cur.lns, ln_t3)
array.push(cur.lbls, lb_t1)
array.push(cur.lbls, lb_t3)
array.push(pat_history, cur)
f_manage_drawings(max_patterns)
if alert_impulse and res_w.pattern_type == 1 and barstate.isconfirmed
alert(str.format("Elliott {0} — Entry: {1} Stop: {2} Target: {3}", res_w.name, str.tostring(res_w.entry_p, format.mintick), str.tostring(res_w.stop_p, format.mintick), str.tostring(res_w.target_p, format.mintick)), alert.freq_once_per_bar_close)
if alert_correction and res_w.pattern_type == 2 and barstate.isconfirmed
alert(str.format("Elliott {0} — Entry: {1} Stop: {2} Target: {3}", res_w.name, str.tostring(res_w.entry_p, format.mintick), str.tostring(res_w.stop_p, format.mintick), str.tostring(res_w.target_p, format.mintick)), alert.freq_once_per_bar_close)
// Dashboard
int dash_piv_total = array.size(pivotHighs) + array.size(pivotLows)
float dash_rr_score = math.min(dash_rr_raw * 2.0, 10.0)
float dash_piv_score = math.min(dash_piv_total * 0.005 * 10.0, 10.0)
string dash_rr_str = f_blockBar(dash_rr_score)
string dash_piv_str = f_blockBar(dash_piv_score)
color dash_rr_col = f_blockColor(dash_rr_score)
color dash_piv_col = f_blockColor(dash_piv_score)
color _bull_c = color.rgb(0, 255, 120)
color _bear_c = color.rgb(255, 60, 80)
color dash_bias_col = dash_bull ? _bull_c : _bear_c
color dash_name_col = dash_bull ? _bull_c : _bear_c
string dash_bias_str = dash_bull ? "▲ Bull" : "▼ Bear"
string dash_type_str = dash_pattern_type == 1 ? "📈 Impulse (12345)" : dash_pattern_type == 2 ? "🔄 Correction (ABC)" : "—"
color dash_type_col = dash_pattern_type == 1 ? col_impulse : dash_pattern_type == 2 ? col_correction : color.rgb(110, 110, 110)
string dash_w1_str = not na(dash_w1_len) ? str.tostring(dash_w1_len, format.mintick) : "—"
string dash_w3_str = not na(dash_w3_len) ? str.tostring(dash_w3_len, format.mintick) : "—"
string dash_w5_str = not na(dash_w5_len) ? str.tostring(dash_w5_len, format.mintick) : "—"
string dash_ratio_str = not na(dash_w1_len) and not na(dash_w3_len) and dash_w1_len > 1e-10 ? str.tostring(math.round(dash_w3_len / dash_w1_len, 3)) + "× W1" : "—"
string dash_w5r_str = not na(dash_w5_ratio) ? str.tostring(math.round(dash_w5_ratio, 3)) + "× W1" : "—"
string dash_ctype_str = dash_corr_type != "" and dash_corr_type != "—" ? dash_corr_type : "—"
string dash_entry_str = not na(dash_entry) ? str.tostring(dash_entry, format.mintick) : "—"
string dash_stop_str = not na(dash_stop) ? str.tostring(dash_stop, format.mintick) : "—"
string dash_target_str= not na(dash_target) ? str.tostring(dash_target, format.mintick) : "—"
string dash_atr_str = str.tostring(atr_val, format.mintick)
string dash_since_str = "⏳ " + str.tostring(dash_bars_since) + " Bars"
string dash_cnt_str = "📈 " + str.tostring(impulse_count) + " 🔄 " + str.tostring(correction_count)
color hdrBg = color.rgb(30, 30, 45, 10)
color altBg = color.rgb(20, 20, 25, 10)
color oddBg = dash_bg_color
var table dash = table.new(position.top_right, 2, 22,
border_width=1, border_color=color.rgb(50, 50, 70, 20),
frame_width=1, frame_color=color.rgb(60, 60, 90, 10))
table.cell(dash, 0, 0, " Elliott Wave Scanner", bgcolor=hdrBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 0, clean_ticker + " · " + timeframe.period, bgcolor=hdrBg, text_color=dash_txt_color, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 1, " Bias", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 1, dash_bias_str, bgcolor=altBg, text_color=dash_bias_col, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 2, " Pattern", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 2, dash_name, bgcolor=oddBg, text_color=dash_name_col, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 3, " Type", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 3, dash_type_str, bgcolor=altBg, text_color=dash_type_col, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 4, " Entry", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 4, dash_entry_str, bgcolor=oddBg, text_color=color.rgb(255, 220, 0), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 5, " Stop", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 5, dash_stop_str, bgcolor=altBg, text_color=color.rgb(255, 60, 80), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 6, " Target", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 6, dash_target_str, bgcolor=oddBg, text_color=color.rgb(0, 255, 160), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 7, " R:R Quality", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 7, dash_rr_str, bgcolor=altBg, text_color=dash_rr_col, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 8, " ─── Impulse ───", bgcolor=hdrBg, text_color=color.new(col_impulse, 30), text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 8, "Waves 1·2·3·4·5", bgcolor=hdrBg, text_color=color.new(col_impulse, 30), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 9, " W1 Length", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 9, dash_w1_str, bgcolor=oddBg, text_color=color.rgb(100, 200, 255), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 10, " W3 Length", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 10, dash_w3_str, bgcolor=altBg, text_color=color.rgb(100, 200, 255), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 11, " W5 Length", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 11, dash_w5_str, bgcolor=oddBg, text_color=color.rgb(100, 200, 255), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 12, " W3/W1 Ratio", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 12, dash_ratio_str, bgcolor=altBg, text_color=color.rgb(180, 255, 80), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 13, " W5/W1 Ratio", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 13, dash_w5r_str, bgcolor=oddBg, text_color=color.rgb(100, 200, 255), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 14, " Corr Type", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 14, dash_ctype_str, bgcolor=altBg, text_color=dash_pattern_type == 2 ? col_correction : color.rgb(110, 110, 110), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 15, " ─── System ───", bgcolor=hdrBg, text_color=color.rgb(80, 80, 100), text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 15, "", bgcolor=hdrBg, text_color=color.rgb(80, 80, 100), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 16, " Pivot Count", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 16, dash_piv_str, bgcolor=oddBg, text_color=dash_piv_col, text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 17, " Bars Since", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 17, dash_since_str, bgcolor=altBg, text_color=color.rgb(140, 180, 255), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 18, " ATR (14)", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 18, dash_atr_str, bgcolor=oddBg, text_color=color.rgb(200, 200, 60), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 19, " Pattern Count", bgcolor=altBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 19, dash_cnt_str, bgcolor=altBg, text_color=color.rgb(180, 180, 200), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 20, use_htf_bias ? " HTF Bias" : "", bgcolor=oddBg, text_color=dash_txt_color, text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 20, use_htf_bias ? (htf_bull ? "▲ Bull" : "▼ Bear") : "", bgcolor=oddBg, text_color=use_htf_bias ? (htf_bull ? _bull_c : _bear_c) : color.rgb(110, 110, 110), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 0, 21, " EWD v2.0", bgcolor=hdrBg, text_color=color.rgb(60, 60, 80), text_halign=text.align_left, text_font_family=font.family_monospace, text_size=size.small)
table.cell(dash, 1, 21, "Pine v6", bgcolor=hdrBg, text_color=color.rgb(60, 60, 80), text_halign=text.align_right, text_font_family=font.family_monospace, text_size=size.small)