statsmodels.tsa.vector_ar.irf.IRAnalysis.plot_cum_effects#

IRAnalysis.plot_cum_effects(orth=False, *, impulse=None, response=None, signif=0.05, plot_params=None, figsize=(10, 10), subplot_params=None, plot_stderr=True, stderr_type='asym', repl=1000, rng=None, err_bands=None)#

Plot cumulative impulse response functions

Parameters:
orthbool, default False

Compute orthogonalized impulse responses

impulse{str, int}

variable providing the impulse

response{str, int}

variable affected by the impulse

signiffloat (0 < signif < 1)

Significance level for error bars, defaults to 95% CI

subplot_paramsdict

To pass to subplot plotting functions. Example: if fonts are too big, pass {‘fontsize’ : 8} or some number to your taste.

plot_paramsdict

Keyword arguments to pass to the individual plotting functions.

figsize(float, float), default (10, 10)

Figure size (width, height in inches)

plot_stderrbool, default True

Plot standard impulse response error bands

stderr_typestr

‘asym’: default, computes asymptotic standard errors ‘mc’: Monte Carlo standard errors (use repl)

replint, default 1000

Number of replications for Monte Carlo standard errors

rng{None, int, array_like[int], numpy.random.Generator, numpy.random.RandomState}, optional

np.random seed for Monte Carlo replications

seed{None, int, array_like[int], numpy.random.Generator, numpy.random.RandomState}, optional

Deprecated since version 0.15: seed has been deprecated. In-line with SPEC-007, use rng for passing a random number generator or seed.

err_bandsndarray of shape (2, periods + 1, neqs, neqs), optional

Pre-computed error bands. The first dimension contains the lower and upper bounds of the confidence interval, respectively. If provided, the internal calculation of standard errors is bypassed and stderr_type is used only for plot formatting.