🔢 Omega Lorentzian Signals – Simple Inputs Guide
Omega Lorentzian Signals is a machine-learning trading indicator that compares current market conditions with similar situations from historical data.
The indicator uses several market features, filters the model predictions, and calculates a probability before displaying Buy or Sell signals.
This guide explains all inputs in a simple and easy-to-read format.
🔢 Installation
🔹 Open MetaTrader 5.
🔹 Go to File → Open Data Folder.
🔹 Open MQL5 → Indicators.
🔹 Copy the .ex5 indicator file into the folder.
🔹 Return to MetaTrader 5.
🔹 Open Navigator (Ctrl+N).
🔹 Right-click Indicators and select Refresh.
🔹 Open the desired symbol and timeframe.
🔹 Drag Omega Lorentzian Signals onto the chart.
🔹 Enter the licence key in InpLicenseKey.
Allow DLL Imports
DLL imports are required for the licence check, Telegram alerts, and WEB / TELEGRAM buttons.
🔹 MT5 DLL Permission — Go to Tools → Options → Expert Advisors and enable Allow DLL imports.
🔹 Indicator DLL Permission — When the indicator settings window opens, go to the Common tab and also enable Allow DLL imports.
Licence Check
🔹 Licence Key [InpLicenseKey] — Enter the licence key on the Inputs tab.
After loading the indicator, check the Experts tab. A message such as “Licence: Licence OK” confirms successful activation.
🔢 General Settings
🔹 Bars to Calculate [MaxBarsCalc] — Default: 3000 — Controls how many recent bars are calculated when the indicator loads. 0 means all available bars. Recommended: 3000 or higher when enough history is available.
For full-quality model calculations, the indicator needs enough historical data for both training and prediction.
The chart should contain at least about 1,050 bars.
Simple Idea: More historical bars provide more data for the model but can increase calculation time.
🔢 Licence
🔹 Licence Key [InpLicenseKey] — Default: (empty) — Enter the licence key from EAFXSTORE.
The key is checked online when the indicator starts and approximately every 30 minutes.
The licence is linked to the trading account.
If the licence server cannot be reached, the indicator normally continues working for a 72-hour grace period after the last successful check.
🔹 Licence Error — Empty Key — If the message says “Please enter your licence key.”, enter the key in InpLicenseKey.
🔹 Licence Error — DLL — If the message says “Please tick Allow DLL imports…”, enable DLL imports.
🔹 Licence Error — Server — If the message says “Cannot reach the licence server.”, check the internet connection or firewall.
🔹 Licence Error — System Clock — If the message says “System clock check failed.”, check the computer date and time.
🔢 Language
🔹 Language [Lang] — Default: LANG_EN — Controls the interface language. Recommended: LANG_EN.
The current build is English only, so changing this setting has no effect.
🔢 Machine-Learning Model
These settings control how the machine-learning model searches historical data and makes predictions.
🔹 Neighbors [Neighbors] — Default: 8 — Controls the number of historical samples used to vote on the prediction. Recommended: 8.
A smaller value such as 4 reacts faster but can be noisier.
A larger value such as 12–16 produces smoother results but reacts more slowly.
When SelfTune is enabled, the model can automatically select 4, 8, or 12.
🔹 Maximum Bars Back [MaxBarsBack] — Default: 2000 — Controls how many previous bars the model searches for similar market conditions. Recommended: 2000.
Higher values provide more historical examples but require more CPU time.
🔹 Sample Step [SampleStep] — Default: 4 — Controls how often the model takes a historical training sample. Recommended: 4.
For example, 2000 / 4 provides approximately 500 training samples.
🔹 Prediction Horizon [Horizon] — Default: 4 — Controls how many bars ahead the model predicts. Recommended: 4.
A bar is classified as Up when price closes higher after the selected horizon.
When DynamicExit is disabled, this value also becomes the holding period.
🔹 Normalization Length [NormLen] — Default: 500 — Controls the window used to scale WaveTrend and CCI values from 0–1. Recommended: 500.
Larger values provide more stable scaling.
Smaller values adapt faster to changing market conditions.
🔹 Maximum History [MaxHistory] — Default: 3000 — Controls how many recent bars are used for predictions, signals, and statistics. Recommended: 3000.
Lower values can make loading faster.
Higher values allow signals and statistics to be displayed further back in history.
Simple Idea: The model looks at historical market conditions and finds similar situations to estimate the next price direction.
🔢 Features
These settings control which market measurements are used by the machine-learning model.
At least one feature should remain enabled. The guide recommends using three or more features.
🔹 Use RSI [UseRSI] — Default: true — Uses RSI 14 to measure recent momentum and the balance between gains and losses. Recommended: true.
🔹 Use WaveTrend [UseWT] — Default: true — Uses WaveTrend to identify overbought and oversold market swings. Recommended: true.
🔹 Use CCI [UseCCI] — Default: true — Uses CCI 20 to measure how far price is from its 20-bar average. Recommended: true.
🔹 Use ADX [UseADX] — Default: true — Uses ADX 20 to measure trend strength. Recommended: true.
ADX measures trend strength rather than trend direction.
🔹 Use Z-Score [UseZ] — Default: true — Uses a 100-bar regression z-score to measure how far price is above or below its linear trend line. Recommended: true.
Simple Idea: These features give the model different information about momentum, trend strength, price position, and market conditions.
🔢 Filters
Filters decide whether a model prediction can become a trading signal.
The dashboard shows the filter status under:
Filters: vol / regime / ADX / kernel
🔹 Volatility Filter [VolFilter] — Default: true — Requires the signal bar's range to be larger than half of ATR(10). Recommended: true.
This helps block signals during very quiet market conditions.
🔹 Regime Filter [RegimeFilter] — Default: true — Checks the direction of the kernel slope before allowing a signal. Recommended: true.
This helps block signals that move against a clearly sloping trend.
🔹 Regime Threshold [RegimeThr] — Default: -0.1 — Controls how much kernel slope is required by the regime filter. Recommended: -0.1.
-0.1 allows signals when the kernel is flat or only slightly against the trade.
0 requires the kernel not to slope against the trade.
Positive values such as 0.05 require the kernel to slope in the same direction as the trade.
🔹 ADX Filter [AdxFilter] — Default: false — Allows signals only when ADX(20) is above 20. Recommended: false by default.
This filter can be useful on lower timeframes and strongly trending instruments.
🔹 Kernel Filter [KernelFilter] — Default: true — Allows Buy signals when the kernel is rising and Sell signals when the kernel is falling. Recommended: true.
🔹 Kernel Bandwidth [KernelH] — Default: 8 — Controls how smooth and responsive the kernel line is. Recommended: 8.
Higher values make the kernel smoother and slower.
Lower values make it react faster.
Typical range: 6–16.
🔹 Kernel Relative Weight [KernelR] — Default: 8.0 — Controls the relative weighting of the rational quadratic kernel. Recommended: 8.0.
Lower values give more weight to older bars.
Higher values focus more on recent bars.
Typical range: 1–25.
🔹 Kernel Calculation Bars [KernelX] — Default: 25 — Controls how many bars are included in the kernel calculation. Recommended: 25.
The calculation uses KernelX + 60 bars, which equals 85 bars with the default setting.
Simple Idea: Filters help remove signals that do not meet the required volatility, trend, ADX, or kernel conditions.
🔢 Self-Tuning & Exit
🔹 Self-Tune [SelfTune] — Default: true — Automatically tests different model configurations and selects the one with the best recent accuracy. Recommended: true.
The model tests 18 configurations using:
K = 4 / 8 / 12
and six weight presets:
Balanced, RSI+, WT+, CCI+, ADX+, Z+
🔹 Tune Every [TuneEvery] — Default: 500 — Controls how often the model performs a new self-tune. Recommended: 500.
Lower values adapt faster but require more CPU and may react more strongly to market noise.
🔹 Tune Window [TuneWindow] — Default: 600 — Controls how many recent bars are used to evaluate each configuration during self-tuning. Recommended: 600.
🔹 Dynamic Exit [DynamicExit] — Default: true — Controls how signals are closed. Recommended: true.
When true, a signal closes when the kernel changes direction.
When false, a signal closes after the selected Horizon.
An opposite signal can also close the current signal in both modes.
🔹 Minimum Probability [MinProbPct] — Default: 60 — Controls the minimum confidence required before a signal is displayed. Recommended: 60.
60 means at least 60% of the selected neighbours must agree with the signal direction.
Higher values such as 65–75 create fewer signals.
Lower values such as 55 create more signals.
Simple Idea: Self-tuning automatically adjusts the model, while MinProbPct controls how much agreement is required before producing a signal.
🔢 Display Settings
🔹 Theme [Theme] — Default: THEME_DARK — Controls the chart colour scheme. Recommended: THEME_DARK.
Available options include:
🔹 THEME_NONE — Keeps the existing chart colours.
🔹 THEME_DARK — Uses the dark theme.
🔹 THEME_MIDNIGHT — Uses a deep-blue theme.
🔹 THEME_LIGHT — Uses a white background with grid.
🔹 Restore Theme [RestoreTheme] — Default: true — Restores the original chart colours when the indicator is removed. Recommended: true.
🔹 Show Kernel [ShowKernel] — Default: true — Displays the kernel regression line on the chart. Recommended: true.
The line is:
🔹 Green — Kernel rising.
🔹 Red — Kernel falling.
🔹 Grey — Kernel flat.
Even when hidden, the kernel is still used by filters and exits.
🔹 Show Probability Label [ShowProbLabel] — Default: true — Shows the signal probability near Buy and Sell arrows. Recommended: true.
🔹 Label Font [LabelFont] — Default: 8 — Controls the probability label font size. Recommended: 8.
🔢 Dashboard
🔹 Show Dashboard [ShowDash] — Default: true — Shows or hides the information panel. Recommended: true.
Hiding the dashboard can slightly reduce CPU usage.
🔹 Dashboard Position [DashRight] — Default: false — Places the dashboard on the left or right side of the chart. Recommended: false.
false = left side.
true = right side.
🔹 Dashboard X Position [DashX] — Default: 10 — Controls the horizontal distance from the selected side of the chart.
🔹 Dashboard Y Position [DashY] — Default: 28 — Controls the vertical distance from the top of the chart.
🔹 Dashboard Scale [DashScale] — Default: 1.5 — Controls the dashboard size. Recommended: 1.0–1.5 for normal screens.
Higher values can be useful on 4K screens.
🔹 Dashboard Font Adjustment [DashFontAdj] — Default: 1.0 — Adjusts the dashboard text size. Recommended: 1.0.
The available range is 0.8–1.3.
The dashboard can also be moved by dragging its header.
🔢 Alerts
Alerts are generated only for live bars and only after a bar closes.
🔹 Popup Alert [AlertPopup] — Default: true — Shows the MetaTrader alert window. Recommended: true.
🔹 Push Alert [AlertPush] — Default: false — Sends notifications to the MT5 mobile application. Recommended: false unless mobile alerts are needed.
MetaQuotes ID must be entered in Tools → Options → Notifications.
🔹 Sound Alert [AlertSoundOn] — Default: false — Plays a sound when an alert is triggered. Recommended: false or true depending on preference.
🔹 Alert Sound File [AlertSoundFile] — Default: alert.wav — Selects the sound file used for alerts. Recommended: alert.wav.
The file must be located in the terminal's Sounds folder.
🔹 Signal Alert [AlertSignal] — Default: true — Sends an alert when a new Buy or Sell signal appears. Recommended: true.
The alert can include direction, probability, price, K, preset, and accuracy.
🔹 Exit Alert [AlertExit] — Default: true — Sends an alert when a signal is closed. Recommended: true.
🔹 Self-Tune Alert [AlertTune] — Default: false — Sends an alert when self-tuning changes K or the weight preset. Recommended: false unless tuning alerts are needed.
🔢 Telegram
🔹 Enable Telegram [TgEnable] — Default: false — Sends enabled indicator alerts to Telegram. Recommended: false unless Telegram alerts are needed.
🔹 Telegram Bot Token [TgToken] — Default: (empty) — Stores the Telegram bot token obtained from @BotFather.
🔹 Telegram Chat ID [TgChatID] — Default: (empty) — Specifies the Telegram chat, group, or channel that receives the alerts.
For channels and supergroups, the ID normally starts with -100.
Telegram Setup
🔹 Step 1 — Open @BotFather in Telegram and send /newbot.
🔹 Step 2 — Copy the bot token into TgToken.
🔹 Step 3 — Open the new bot and press Start, or add it to the required group or channel.
🔹 Step 4 — Find the required Chat ID and enter it in TgChatID.
🔹 Step 5 — Set TgEnable = true.
🔹 Step 6 — Make sure Allow DLL imports is enabled.
🔹 Step 7 — Click TEST TG on the dashboard.
A successful test shows a confirmation message.
🔢 About
🔹 Website [AboutWeb] — Default: https://eafxstore.com — Opens the website from the dashboard's WEB button.
🔹 Telegram [AboutTelegram] — Default: https://t.me/eafxstore6789 — Opens the support Telegram link from the dashboard's TELEGRAM button.
🔢 Quick Troubleshooting
🔹 Nothing appears on the chart — Check the licence, enable DLL imports, and make sure the chart has at least about 1,050 bars of history.
🔹 Dashboard shows “not yet (needs history)” — The indicator does not have enough historical data for self-tuning. Load more history or increase MaxBarsCalc.
🔹 Very few signals — Lower MinProbPct, turn off AdxFilter, lower RegimeThr, or check which filter shows × on the dashboard.
🔹 Too many signals — Raise MinProbPct, enable AdxFilter, or set RegimeThr = 0.
🔹 MT5 feels slow — Reduce MaxBarsCalc, MaxHistory, and MaxBarsBack, or increase SampleStep.
🔹 Telegram alerts are not received — Check TgToken, TgChatID, bot status, DLL imports, and internet access.
🔹 Push notifications are not received — Check the MetaQuotes ID and make sure push notifications are enabled in Tools → Options → Notifications.
🔢 Recommended Starting Settings
For a simple starting configuration, the default settings can be used:
🔹 Bars to Calculate [MaxBarsCalc] — 3000
🔹 Licence Key [InpLicenseKey] — Enter the valid licence key.
🔹 Language [Lang] — LANG_EN
🔹 Neighbors [Neighbors] — 8
🔹 Maximum Bars Back [MaxBarsBack] — 2000
🔹 Sample Step [SampleStep] — 4
🔹 Prediction Horizon [Horizon] — 4
🔹 Normalization Length [NormLen] — 500
🔹 Maximum History [MaxHistory] — 3000
🔹 Use RSI [UseRSI] — true
🔹 Use WaveTrend [UseWT] — true
🔹 Use CCI [UseCCI] — true
🔹 Use ADX [UseADX] — true
🔹 Use Z-Score [UseZ] — true
🔹 Volatility Filter [VolFilter] — true
🔹 Regime Filter [RegimeFilter] — true
🔹 Regime Threshold [RegimeThr] — -0.1
🔹 ADX Filter [AdxFilter] — false
🔹 Kernel Filter [KernelFilter] — true
🔹 Kernel Bandwidth [KernelH] — 8
🔹 Kernel Relative Weight [KernelR] — 8.0
🔹 Kernel Calculation Bars [KernelX] — 25
🔹 Self-Tune [SelfTune] — true
🔹 Tune Every [TuneEvery] — 500
🔹 Tune Window [TuneWindow] — 600
🔹 Dynamic Exit [DynamicExit] — true
🔹 Minimum Probability [MinProbPct] — 60
🔹 Theme [Theme] — THEME_DARK
🔹 Restore Theme [RestoreTheme] — true
🔹 Show Kernel [ShowKernel] — true
🔹 Show Probability Label [ShowProbLabel] — true
🔹 Label Font [LabelFont] — 8
🔹 Show Dashboard [ShowDash] — true
🔹 Dashboard Position [DashRight] — false
🔹 Dashboard X [DashX] — 10
🔹 Dashboard Y [DashY] — 28
🔹 Dashboard Scale [DashScale] — 1.5
🔹 Dashboard Font Adjustment [DashFontAdj] — 1.0
🔹 Popup Alert [AlertPopup] — true
🔹 Push Alert [AlertPush] — false
🔹 Sound Alert [AlertSoundOn] — false
🔹 Signal Alert [AlertSignal] — true
🔹 Exit Alert [AlertExit] — true
🔹 Self-Tune Alert [AlertTune] — false
🔹 Enable Telegram [TgEnable] — false
🔢 Simple Omega Lorentzian Signals Logic
The indicator can be understood through a simple process:
🔹 Step 1 — Collect Market Data — The indicator reads the selected historical market information.
🔹 Step 2 — Calculate Features — RSI, WaveTrend, CCI, ADX, and Z-Score provide different measurements of market conditions.
🔹 Step 3 — Search Historical Data — The model searches for historical situations that are similar to the current market.
🔹 Step 4 — Find Neighbours — The selected Neighbors value determines how many similar historical samples participate in the prediction.
🔹 Step 5 — Calculate Probability — The model checks how many neighbours agree on the expected direction.
🔹 Step 6 — Apply Filters — Volatility, regime, ADX, and kernel filters decide whether the prediction can become a signal.
🔹 Step 7 — Check Minimum Probability — The signal must meet the MinProbPct requirement.
🔹 Step 8 — Display Buy or Sell Signal — When the required conditions are satisfied, the indicator displays the signal and probability.
🔹 Step 9 — Manage the Exit — With DynamicExit = true, the signal can close when the kernel changes direction. With DynamicExit = false, the signal closes after the selected Horizon.
🔹 Step 10 — Self-Tune — When SelfTune = true, the model periodically tests different configurations and selects the one with the best recent accuracy.
🔢 Simple Summary
Omega Lorentzian Signals combines machine learning, historical pattern matching, technical features, filters, probability analysis, and automatic self-tuning.
The basic concept is:
Current Market → Compare With Historical Data → Find Similar Neighbours → Calculate Probability → Apply Filters → Generate Signal
The main settings that control signal frequency are:
🔹 Neighbors [Neighbors] — Controls how many historical samples vote.
🔹 Horizon [Horizon] — Controls how far ahead the model predicts.
🔹 MinProbPct [MinProbPct] — Controls the minimum confidence required.
🔹 RegimeThr [RegimeThr] — Controls the trend condition.
🔹 AdxFilter [AdxFilter] — Controls whether ADX must confirm a trend.
🔹 KernelFilter [KernelFilter] — Controls whether the kernel direction must agree with the signal.
🔹 SelfTune [SelfTune] — Controls automatic model optimisation.
