NinjaScript tutorials from someone who’s been writing it since NT6.5. Start at the top if you’re new — or jump to whatever you’re stuck on.
The prior post let you stack any number of hidden layers — but every weight update was still plain stochastic gradient descent: one update per bar, one fixed step size for every weight. Plain SGD is the simplest possible training rule, and it works on shallow networks, but two things tend to go wrong at…
Tooltips live in two places in a NinjaScript indicator. The first is the settings dialog — the small hover text that explains what a property does. The second is the chart itself — labels and annotations attached to drawn objects that show context when a trader hovers over them. Both are cheap to add, both…
The prior post built a single-hidden-layer neural network — three inputs, six tanh neurons in one hidden layer, one sigmoid output, 31 parameters. It worked because the hidden layer can compose the raw features into non-linear combinations that a single neuron cannot represent. The post closed with a hint: “add a hidden layer + backprop…
A polished indicator hides settings that don’t apply to the current state. When a ShowLabels bool is off, the LabelColor, LabelSize, and LabelFont properties should disappear from the grid — they’re irrelevant unless labels are enabled. Standard NinjaScript attributes don’t give you that directly. There’s no [Display(EnabledWhen = “ShowLabels”)]. Three workarounds do it. Each has…
The prior post built a single-neuron model — three weights, a bias, and a sigmoid that turned a weighted sum into P(up). It worked, but its limits were architectural: a single neuron can only carve the feature space with a straight line. Any pattern that requires the features to interact non-linearly — “high distFromMa AND…
A three-property indicator fits fine in NinjaTrader’s default “Parameters” bucket. A fifteen-property indicator doesn’t. Once an indicator has ATR inputs, confirmation thresholds, display toggles, color choices, and alert settings, the property grid needs real structure — otherwise users scroll past the setting they actually wanted to change. Last week’s post introduced [Display(GroupName, Order)] and [Gui.CategoryOrder]….
The k-Nearest Neighbors post introduced one half of supervised learning: memorizing models that hold onto every historical bar and search through them at prediction time. This post is the other half. Instead of remembering every example, an online-learning model carries a small set of weights — three numbers, plus a bias term, in this case…
A property’s C# identifier isn’t always what you want users to see in the indicator settings dialog. Period reads fine; atrPeriodLookback looks like a raw variable name. The [Display] attribute lets you set a friendlier label, add a tooltip, group related properties together, and control the order they appear. Four parameters cover almost all of…
Search TradingView’s community scripts for terms like “machine learning” or “AI” and you’ll find a long list of community-built indicators that lean on one of the simplest and oldest algorithms in ML. k-Nearest Neighbors — the algorithm we’re building from scratch in NinjaScript today — dates to the 1950s and remains one of the most…
You wrote an indicator. It works great on a chart. Now you want to use it from a strategy — you write var sma = MySma(20); in State.DataLoaded and expect the cross-script reference to just work. Or you want to optimize Period in Strategy Analyzer and see which value produces the best backtest. How does…
The Account Copier mirrors one leader account to every follower in lockstep, on the same machine, across your LAN, or through the cloud relay. Free on one account.
Notifications
