# Technical analysis

> The ta.* namespace: averages, oscillators, bands, crosses, pivots and multi-value studies, with working recipes.

`ta.*` holds the building blocks: more than seventy functions, from `ta.sma` to `ta.supertrend`. They all follow the bar-by-bar model, so each returns a series and handles its own warm-up (the first bars are `na`).

## Averages

| Function | Notes |
| --- | --- |
| `ta.sma(src, len)` | Simple average |
| `ta.ema(src, len)` | Exponential, reacts faster |
| `ta.wma(src, len)` | Linearly weighted |
| `ta.hma(src, len)` | Hull, smooth and fast |
| `ta.rma(src, len)` | Wilder's smoothing, used inside RSI and ATR |
| `ta.vwma(src, len)` | Volume weighted (see the note on volume below) |
| `ta.alma(src, len, offset, sigma)` | Arnaud Legoux |

## Oscillators and volatility

`ta.rsi`, `ta.stoch`, `ta.cci`, `ta.mfi`, `ta.mom`, `ta.roc`, `ta.tsi`, `ta.wpr`, `ta.atr`, `ta.tr`, `ta.stdev`, `ta.dev`, and more.

```pine
//@version=6
indicator("Stochastic")
k = ta.sma(ta.stoch(close, high, low, 14), 3)
d = ta.sma(k, 3)
plot(k, "%K", color = color.new(#8b7cff, 0), linewidth = 2)
plot(d, "%D", color = color.new(#ffb547, 0))
hline(80, "Upper", color = color.gray)
hline(20, "Lower", color = color.gray)
```

## Studies that return several values

Some functions return a tuple. Destructure it with square brackets:

```pine
//@version=6
indicator("MACD")
[macdLine, signalLine, hist] = ta.macd(close, 12, 26, 9)
plot(hist, "Histogram", style = plot.style_columns, color = hist >= 0 ? color.new(color.teal, 40) : color.new(color.red, 40))
plot(macdLine, "MACD", color = color.new(#8b7cff, 0))
plot(signalLine, "Signal", color = color.new(#ffb547, 0))
```

Others in this family: `ta.bb` (basis, upper, lower), `ta.kc` (Keltner), `ta.dmi` (+DI, -DI, ADX), `ta.supertrend` (value, direction).

```pine
//@version=6
indicator("Supertrend", overlay = true)
[st, dir] = ta.supertrend(3, 10)
plot(dir < 0 ? st : na, "Up trend", color = color.teal, style = plot.style_linebr, linewidth = 2)
plot(dir > 0 ? st : na, "Down trend", color = color.red, style = plot.style_linebr, linewidth = 2)
```

In `ta.supertrend`, a **negative** direction means an up trend.

## Crosses, highs and lows, pivots

- `ta.crossover(a, b)`, `ta.crossunder(a, b)`, `ta.cross(a, b)`: `true` on the bar the cross happens.
- `ta.highest(src, len)`, `ta.lowest(src, len)`, and `ta.highestbars` / `ta.lowestbars` for how many bars ago.
- `ta.pivothigh(src, left, right)` and `ta.pivotlow(...)`: the pivot price, reported `right` bars after the pivot (because it needs those bars to confirm). `na` elsewhere.
- `ta.valuewhen(cond, src, n)`: the value of `src` the n-th last time `cond` was true.
- `ta.barssince(cond)`: bars since `cond` was last true.

```pine
//@version=6
indicator("Pivot levels", overlay = true)
ph = ta.pivothigh(high, 10, 10)
pl = ta.pivotlow(low, 10, 10)
plot(ta.valuewhen(not na(ph), ph, 0), "Last pivot high", color = color.red, style = plot.style_stepline)
plot(ta.valuewhen(not na(pl), pl, 0), "Last pivot low", color = color.teal, style = plot.style_stepline)
plotshape(not na(ph), "PH", style = shape.triangledown, location = location.abovebar, color = color.red, offset = -10, size = size.tiny)
plotshape(not na(pl), "PL", style = shape.triangleup, location = location.belowbar, color = color.teal, offset = -10, size = size.tiny)
```

`offset = -10` draws each marker back on the bar that actually was the pivot.

## A note on volume

Volume on this feed is a **tick count** per bar, not traded size. Indicators that compare volume between bars (`ta.obv`, `ta.mfi`, plotting volume itself) are meaningful on it. Ones that treat it as size, such as `ta.vwap` and accumulation/distribution, still compute, and the editor warns that the result is approximate. See [The data](/docs/data).
