Comparing multiple tickers¶

Overlay another ticker's close price on the primary chart to compare relative performance. Merging price series is ordinary data preparation — a join plus an optional rebase — so it is done inline here, in each backend, rather than through a library helper. With both series scaled to the same starting level, relative performance reads directly off the chart.

In [1]:

import yfinance as yf

from mplchart.chart import Chart
from mplchart.primitives import Candlesticks, Volume, LinePlot
from mplchart.utils import normalize_prices

aapl = normalize_prices(yf.Ticker("AAPL").history(period="5y"))
msft = normalize_prices(yf.Ticker("MSFT").history(period="5y"))
aapl.tail(3)

import yfinance as yf from mplchart.chart import Chart from mplchart.primitives import Candlesticks, Volume, LinePlot from mplchart.utils import normalize_prices aapl = normalize_prices(yf.Ticker("AAPL").history(period="5y")) msft = normalize_prices(yf.Ticker("MSFT").history(period="5y")) aapl.tail(3)

Out[1]:

	open	high	low	close	volume	dividends	stock splits
date
2026-07-22 00:00:00-04:00	327.869995	329.000000	323.339996	325.890015	38755900	0.0	0.0
2026-07-23 00:00:00-04:00	321.730011	323.299988	319.350006	321.660004	40840800	0.0	0.0
2026-07-24 00:00:00-04:00	321.790009	334.369995	321.619995	333.019989	47443900	0.0	0.0

Merging with pandas¶

Align the secondary close on the primary index, scale it to the primary's close at the first common date, and assign it as a named column.

In [2]:

msft_close = msft["close"].reindex(aapl.index)
first = msft_close.first_valid_index()
scale = aapl.loc[first, "close"] / msft_close[first]

prices = aapl.assign(msft=msft_close * scale)

Chart(prices, title="AAPL vs MSFT (rebased)", max_bars=250).plot(
    Candlesticks(), LinePlot("msft"), Volume(),
).show()

msft_close = msft["close"].reindex(aapl.index) first = msft_close.first_valid_index() scale = aapl.loc[first, "close"] / msft_close[first] prices = aapl.assign(msft=msft_close * scale) Chart(prices, title="AAPL vs MSFT (rebased)", max_bars=250).plot( Candlesticks(), LinePlot("msft"), Volume(), ).show()

Merging with polars¶

The same merge in polars: join on the date column, rebase with expressions.

In [3]:

import polars as pl

pl_aapl = pl.from_pandas(aapl.reset_index(names="date"))
pl_msft = pl.from_pandas(msft.reset_index(names="date")).select("date", msft="close")

merged = pl_aapl.join(pl_msft, on="date", how="left")
scale = merged.select((pl.col("close") / pl.col("msft")).drop_nulls().first()).item()
merged = merged.with_columns(pl.col("msft") * scale)

Chart(merged, title="AAPL vs MSFT (rebased, polars)", max_bars=250).plot(
    Candlesticks(), LinePlot("msft"), Volume(),
).show()

import polars as pl pl_aapl = pl.from_pandas(aapl.reset_index(names="date")) pl_msft = pl.from_pandas(msft.reset_index(names="date")).select("date", msft="close") merged = pl_aapl.join(pl_msft, on="date", how="left") scale = merged.select((pl.col("close") / pl.col("msft")).drop_nulls().first()).item() merged = merged.with_columns(pl.col("msft") * scale) Chart(merged, title="AAPL vs MSFT (rebased, polars)", max_bars=250).plot( Candlesticks(), LinePlot("msft"), Volume(), ).show()

The merge is deliberately transparent: close-only, rebased at the first common date. Those are decisions you may want to adapt — resampling mismatched calendars, currency normalization, rebasing at a chosen date — which is why it lives in your code, not the library.