Using Indicators¶

pandas only

The mintalib.indicators module provides composable indicator objects that bind a calculation with its parameters (pandas only — for polars, use mintalib.expressions).

Indicators are named in upper case (e.g. SMA, EMA, MACD). An indicator instance is callable and can be passed directly to prices.assign() or invoked as SMA(50)(prices). The | operator chains indicators: EMA(20) | ROC(1) means ROC applied after EMA.

In [4]:

Copied!

import pandas as pd

from mintalib.samples import sample_prices
from mintalib.indicators import EMA, SMA, ROC, RSI, LOG, BBANDS, MACD

prices = sample_prices()
import pandas as pd

from mintalib.samples import sample_prices
from mintalib.indicators import EMA, SMA, ROC, RSI, LOG, BBANDS, MACD

prices = sample_prices()

Basic Usage¶

An indicator instance is a callable. Applied to a DataFrame, series-based indicators use the close column by default — the item parameter selects another column. A pandas Series or numpy array can be passed directly as well (results always come back as pandas objects):

In [5]:

Copied!

prices.pipe(SMA(50))

prices.pipe(SMA(50))

Out[5]:

date
1980-12-12           NaN
1980-12-15           NaN
1980-12-16           NaN
1980-12-17           NaN
1980-12-18           NaN
                 ...    
2026-07-31    309.499400
2026-08-03    309.522800
2026-08-04    309.610601
2026-08-05    309.654200
2026-08-06    309.772801
Length: 11504, dtype: float64

In [6]:

Copied!

prices.pipe(RSI(14))
prices.pipe(RSI(14))

Out[6]:

date
1980-12-12          NaN
1980-12-15          NaN
1980-12-16          NaN
1980-12-17          NaN
1980-12-18          NaN
                ...    
2026-07-31    43.245950
2026-08-03    40.340781
2026-08-04    44.685147
2026-08-05    45.839619
2026-08-06    48.183314
Length: 11504, dtype: float64

Chaining¶

The | operator chains indicators left to right: LOG() | EMA(20) | ROC(1) applies LOG first, then EMA, then ROC.

In [7]:

Copied!

prices.assign(
    trend=LOG() | EMA(20) | ROC(1)
)
prices.assign(
    trend=LOG() | EMA(20) | ROC(1)
)

Out[7]:

	open	high	low	close	volume	trend
date
1980-12-12	0.098207	0.098634	0.098207	0.098207	469033600	NaN
1980-12-15	0.093510	0.093510	0.093083	0.093083	175884800	NaN
1980-12-16	0.086678	0.086678	0.086251	0.086251	105728000	NaN
1980-12-17	0.088386	0.088813	0.088386	0.088386	86441600	NaN
1980-12-18	0.090949	0.091376	0.090949	0.090949	73449600	NaN
...	...	...	...	...	...	...
2026-07-31	304.809998	310.690002	300.000000	308.910004	132489100	-0.077130
2026-08-03	309.579987	311.799988	302.559998	303.420013	75052000	-0.099409
2026-08-04	302.730011	310.420013	301.320007	309.380005	68001000	-0.057922
2026-08-05	309.359985	311.709991	305.670013	311.000000	49178700	-0.043810
2026-08-06	314.339996	316.289398	313.489990	314.260010	7165536	-0.022424

11504 rows × 6 columns

The Assign Idiom¶

Because indicators are callables, they can be passed directly to prices.assign, which invokes each with the DataFrame. Since assign processes keyword arguments sequentially, pd.col expressions can reference columns created earlier in the same call:

In [8]:

Copied!





prices.assign(
    sma50 = SMA(50),
    sma200 = SMA(200),
    rsi = RSI(14),
    slope = LOG() | EMA(20) | ROC(1),
    uptrend = pd.col("sma50") > pd.col("sma200")
).iloc[:, -5:]
prices.assign(
    sma50 = SMA(50),
    sma200 = SMA(200),
    rsi = RSI(14),
    slope = LOG() | EMA(20) | ROC(1),
    uptrend = pd.col("sma50") > pd.col("sma200")
).iloc[:, -5:]

Out[8]:

	sma50	sma200	rsi	slope	uptrend
date
1980-12-12	NaN	NaN	NaN	NaN	False
1980-12-15	NaN	NaN	NaN	NaN	False
1980-12-16	NaN	NaN	NaN	NaN	False
1980-12-17	NaN	NaN	NaN	NaN	False
1980-12-18	NaN	NaN	NaN	NaN	False
...	...	...	...	...	...
2026-07-31	309.499400	277.661274	43.245950	-0.077130	True
2026-08-03	309.522800	277.943019	40.340781	-0.099409	True
2026-08-04	309.610601	278.246737	44.685147	-0.057922	True
2026-08-05	309.654200	278.567977	45.839619	-0.043810	True
2026-08-06	309.772801	278.881386	48.183314	-0.022424	True

11504 rows × 5 columns

Multi-Output Indicators¶

Multi-output indicators return a DataFrame, so they cannot be assigned to a single column — join the result instead:

In [9]:

Copied!

prices.pipe(BBANDS(20))
prices.pipe(BBANDS(20))

Out[9]:

	upperband	middleband	lowerband
date
1980-12-12	NaN	NaN	NaN
1980-12-15	NaN	NaN	NaN
1980-12-16	NaN	NaN	NaN
1980-12-17	NaN	NaN	NaN
1980-12-18	NaN	NaN	NaN
...	...	...	...
2026-07-31	344.002618	324.3670	304.731382
2026-08-03	345.001166	323.9050	302.808834
2026-08-04	345.104607	323.8410	302.577394
2026-08-05	345.243987	323.7215	302.199012
2026-08-06	345.299028	323.6235	301.947972

11504 rows × 3 columns

Pandas Expressions¶

With pandas >= 3.0, as_expr() converts an indicator into a pandas Expression. For multi-output indicators, as_expr(item) picks a single output — which makes them usable inside assign after all:

In [10]:

Copied!





prices.assign(
    oversold=RSI(14).as_expr()<30,
    overbought=RSI(14).as_expr()>70,
)
prices.assign(
    oversold=RSI(14).as_expr()<30,
    overbought=RSI(14).as_expr()>70,
)

Out[10]:

	open	high	low	close	volume	oversold	overbought
date
1980-12-12	0.098207	0.098634	0.098207	0.098207	469033600	False	False
1980-12-15	0.093510	0.093510	0.093083	0.093083	175884800	False	False
1980-12-16	0.086678	0.086678	0.086251	0.086251	105728000	False	False
1980-12-17	0.088386	0.088813	0.088386	0.088386	86441600	False	False
1980-12-18	0.090949	0.091376	0.090949	0.090949	73449600	False	False
...	...	...	...	...	...	...	...
2026-07-31	304.809998	310.690002	300.000000	308.910004	132489100	False	False
2026-08-03	309.579987	311.799988	302.559998	303.420013	75052000	False	False
2026-08-04	302.730011	310.420013	301.320007	309.380005	68001000	False	False
2026-08-05	309.359985	311.709991	305.670013	311.000000	49178700	False	False
2026-08-06	314.339996	316.289398	313.489990	314.260010	7165536	False	False

11504 rows × 7 columns

Studies over multiple symbols¶

An IndicatorBundle collects several indicators into a reusable study. Apply it to each symbol independently with pandas groupby.apply.

In [ ]:

Copied!

from mintalib.indicators import IndicatorBundle, MACD, SMA

dataset = pd.concat(
    [prices.assign(symbol=symbol) for symbol in ["AAA", "BBB", "CCC"]]
)

study = IndicatorBundle(MACD(), sma20=SMA(20), sma50=SMA(50))

dataset.groupby("symbol")[prices.columns].apply(study)
from mintalib.indicators import IndicatorBundle, MACD, SMA

dataset = pd.concat(
    [prices.assign(symbol=symbol) for symbol in ["AAA", "BBB", "CCC"]]
)

study = IndicatorBundle(MACD(), sma20=SMA(20), sma50=SMA(50))

dataset.groupby("symbol")[prices.columns].apply(study)

Out[ ]:

		macd	macdsignal	macdhist	sma20	sma50
symbol	date
AAA	1980-12-12	NaN	NaN	NaN	NaN	NaN
	1980-12-15	NaN	NaN	NaN	NaN	NaN
	1980-12-16	NaN	NaN	NaN	NaN	NaN
	1980-12-17	NaN	NaN	NaN	NaN	NaN
	1980-12-18	NaN	NaN	NaN	NaN	NaN
...	...	...	...	...	...	...
CCC	2026-07-31	6.894136	8.260785	-1.366649	324.3670	309.499400
	2026-08-03	4.536518	7.515932	-2.979414	323.9050	309.522800
	2026-08-04	3.113124	6.635370	-3.522246	323.8410	309.610601
	2026-08-05	2.091683	5.726633	-3.634950	323.7215	309.654200
	2026-08-06	1.527629	4.886832	-3.359203	323.6235	309.772801

34512 rows × 5 columns

In [ ]: