gs_quant.timeseries.analysis.lag

lag(x, obs=1, mode=LagMode.EXTEND)

Lag timeseries by a number of observations or a relative date.

Parameters:

• x (Series) – timeseries of prices

• obs (Union[Window, int, str]) – non-zero integer (number of observations) or relative date e.g. “-90d”, “1d”, “1m”, “1y”

• mode (LagMode) – whether to extend series index (into the future)

Return type:

Series

Returns:

date-based time series of return

Usage

Shift the series backwards by a specified number of observations:

\(R_t = X_{t-obs}\)

where \(obs\) is the number of observations to lag series

Examples

Lag series by 2 observations:

>>> prices = generate_series(100)
>>> lagged = lag(prices, 2)

Lag series by 1 year:

>>> prices = generate_series(100)
>>> lagged = lag(prices, '1y')

See also

diff()