auto.arima（）在启用并行化时输出的信息似乎与不使用并行化时输出的信息不同。具体而言，启用并行化时，输出不会指定p，d和q的值，也不会报告AICc或BIC信息标准。
没有并行化的示例：
> model <- auto.arima(ts,xreg=tax.ts,ic=c("bic"),stepwise=FALSE,approximation=FALSE)
Series: ts
ARIMA(0,1,1)
Coefficients:
ma1 tax
0.4589 -0.7035
s.e. 0.0806 0.6969
sigma^2 estimated as 0.6028: log likelihood=-138.36
AIC=282.72 AICc=282.92 BIC=291.08
使用paralellization：
> model <- auto.arima(ts,xreg=tax.ts,ic=c("bic"),stepwise=FALSE,approximation=FALSE,parallel=TRUE,num.cores=12)
Call:
auto.arima(x = structure(list(rwhitebread = c(89.2749786376953, 89.2966079711914,
88.9023666381836, 88.6773300170898, 89.0641860961914, 89.6114883422852, 89.5789642333984,
89.3341522216797, 89.0179290771484, 88.4036026000977, 88.309211730957, 87.6592483520508,
87.3339614868164, 87.2614440917969, 87.7638092041016, 87.3240356445312, 88.6289749145508,
90.0923461914062, 90.2542266845703, 90.6454772949219, 90.7220764160156, 90.4558944702148,
90.8726654052734, 91.7483978271484, 92.7035140991211, 92.5734100341797, 92.4802780151367,
92.6397476196289, 93.3086471557617, 92.8956298828125, 92.6798477172852, 94.3108444213867,
96.7846298217773, 99.8294296264648, 100.150718688965, 103.013710021973, 103.426765441895,
102.606864929199, 102.679161071777, 103.586181640625, 106.069953918457, 105.177772521973,
104.857940673828, 104.737091064453, 104.414154052734, 103.719192504883, 101.936988830566,
101.443199157715, 101.572723388672, 101.843193054199, 101.829467773438, 101.567581176758,
102.02978515625, 101.891319274902, 101.730346679688, 101.729293823242, 101.258888244629,
100.753799438477, 100.468338012695, 100.157859802246, 99.6511154174805, 99.5886764526367,
98.7140121459961, 98.9120025634766, 99.6159439086914, 100.186874389648, 99.477912902832,
99.2161636352539, 99.1324768066406, 99.4329071044922, 99.8656692504883, 100.098243713379,
100.050888061523, 100.484085083008, 100.259925842285, 101.157867431641, 102.265960693359,
102.379020690918, 102.103042602539, 102.0927734375, 101.972709655762, 101.002868652344,
99.5803604125977, 99.2621994018555, 100.411376953125, 101.165588378906, 101.440376281738,
101.992752075195, 102.306083679199, 102.090476989746, 101.909233093262, 105.202362060547,
108.684272766113, 108.776985168457, 108.79345703125, 108.924339294434, 108.688865661621,
108.240608215332, 107.647567749023, 106.971420288086, 107.167625427246, 106.995262145996,
107.178039550781, 107.176040649414, 106.977165222168, 106.254096984863, 104.939926147461,
104.52613067627, 104.082489013672, 104.322189331055, 103.974395751953, 104.827789306641,
105.576438903809, 105.718490600586, 105.364852905273, 105.030937194824, 104.75341796875,
103.981002807617, 103.218376159668, 102.868370056152, 102.812316894531)), .Names = "rwhitebread", row.names = c(NA,
-121L), class = "data.frame"), ic = c("bic"), stepwise = FALSE, approximation = FALSE,
xreg = tax.ts, parallel = TRUE, num.cores = 12)
Coefficients:
ma1 tax
0.4589 -0.7035
s.e. 0.0806 0.6969
sigma^2 estimated as 0.6028: log likelihood = -138.36, aic = 282.72
有没有办法从并行命令获取相同的输出，或以其他方式提取模型和AICc / BIC中使用的参数？