如何将回归应用于 R 中的动物园数据框？

Question

我有一个动物园数据框，其中一部分看起来像：

                TLT      PCY      SHY
2015-04-15 122.1545 28.16594 84.63730
2015-04-16 123.9836 28.11196 84.72825
2015-04-17 124.4341 28.07958 84.70804
2015-04-20 125.2258 28.15514 84.74845
2015-04-21 125.9629 28.18753 84.76866
2015-04-22 126.1267 28.28469 84.79897
2015-04-23 126.4923 28.30628 84.85056
2015-04-24 124.5478 28.34947 84.81010
2015-04-27 123.3290 28.36026 84.63820
2015-04-28 124.3560 28.38185 84.83033
2015-04-29 123.5618 28.45742 84.76965
2015-04-30 122.8086 28.53299 84.73932
2015-05-01 124.0000 28.57618 84.80000

或者看跌期权看起来像：

structure(c(122.154509543852, 123.983621421486, 124.434078860684, 
    125.225787705634, 125.962885778852, 126.126687074014, 126.492320227327, 
    124.547751779181, 123.328974601471, 124.356047962478, 123.561779964634, 
    122.808613513252, 124, 28.1659442697104, 28.1119613477074, 28.079579550721, 
    28.1551370770225, 28.187532134368, 28.2846907856862, 28.3062786503438, 
    28.3494676400181, 28.3602615723469, 28.3818494370044, 28.4574202236651, 
    28.5329910103257, 28.57618, 84.6373035627333, 84.7282463827038, 
    84.7080380110033, 84.7484547544043, 84.7686631261047, 84.7989705363372, 
    84.8505569604793, 84.8100990385319, 84.6381991961199, 84.8303279995055, 
    84.769651411221, 84.7393234117153, 84.8), .Dim = c(13L, 3L), .Dimnames = 
    list(NULL, c("TLT", "PCY", "SHY")), index = structure(c(16540, 
16541, 16542, 16545, 16546, 16547, 16548, 16549, 16552, 16553, 
16554, 16555, 16556), class = "Date"), class = "zoo")

我想对该数据框的列应用线性回归（或一阶多项式拟合），并将结果作为一个向量，向量中的值是每列拟合线的斜率数据。有没有一种方法可以使用“apply”和匿名函数来使用“lm”执行此操作，或者是否有库函数可以执行此操作？

Answer 1

尝试

 apply(zz, 2, function(q) coef(lm(q ~ time(q)))[2])

如何将回归应用于 R 中的动物园数据框？

How to apply regression to a zoo data frame in R?

r

curve-fitting

zoo