1.读取数据
jj = scan("http://www.stat.pitt.edu/stoffer/tsa2/data/jj.dat")
jj <- scan("http://www.stat.pitt.edu/stoffer/tsa2/data/jj.dat")
scan("http://www.stat.pitt.edu/stoffer/tsa2/data/jj.dat") -> jj
> jj<-scan("http://www.stat.pitt.edu/stoffer/tsa2/data/jj.dat") Read 84 items > jj [1] 0.710000 0.630000 0.850000 0.440000 0.610000 0.690000 [7] 0.920000 0.550000 0.720000 0.770000 0.920000 0.600000 [13] 0.830000 0.800000 1.000000 0.770000 0.920000 1.000000 [19] 1.240000 1.000000 1.160000 1.300000 1.450000 1.250000 [25] 1.260000 1.380000 1.860000 1.560000 1.530000 1.590000 [31] 1.830000 1.860000 1.530000 2.070000 2.340000 2.250000 [37] 2.160000 2.430000 2.700000 2.250000 2.790000 3.420000 [43] 3.690000 3.600000 3.600000 4.320000 4.320000 4.050000 [49] 4.860000 5.040000 5.040000 4.410000 5.580000 5.850000 [55] 6.570000 5.310000 6.030000 6.390000 6.930000 5.850000 [61] 6.930000 7.740000 7.830000 6.120000 7.740000 8.910000 [67] 8.280000 6.840000 9.540000 10.260000 9.540000 8.729999 [73] 11.880000 12.060000 12.150000 8.910000 14.040000 12.960000 [79] 14.850000 9.990000 16.200000 14.670000 16.020000 11.610000
scan
scan读入的数据生成向量类型
向量
1.基本元素为：数值（numeric）、字符（character）、逻辑值（logical）、复数型（complex）
2.向量不需要定义类型，可直接赋值。
生成一个空向量；x<-c();
给向量赋值。x<-c(0,1,2,3);
3.向量的元素下标取值是以1开始
4.如果一个向量中有一个字符，则该向量的类型会变成字符.mode(jj)
5.如果逻辑变量与数值在一起，则为转换成数值。TRUE转变成1 and FALSE 转变成 0
> mode(jj) [1] "numeric" > test<-c(1,2,'a') > mode(test) [1] "character" > test1<-c(1,2,true) 错误: 找不到对象'true' > test1<-c(1,2,TRUE) > mode(test1) [1] "numeric"
6.在R语言中没有单一的整数、单一字符的概念. X<-2；X<-'a'；R都是当作向量来处理，只是这个向量只包括单一值.
7.给向量各元素命名: names(x)
> demo<-1:3 > fix(demo) > names(demo)<-c('a','b','c','d') 错误于names(demo) <- c("a", "b", "c", "d") : 'names'属性的长度[4]必需和矢量的长度[3]一样 > names(demo)<-c('a','b','c') > demo a b c 1 2 3 > names(demo)<-c('d','e','f') > demo d e f 1 2 3
jj转变为一个时间序列对象
> jj = ts(jj, start=1960, frequency=4)
> jj
Qtr1 Qtr2 Qtr3 Qtr4
1960 0.710000 0.630000 0.850000 0.440000
1961 0.610000 0.690000 0.920000 0.550000
1962 0.720000 0.770000 0.920000 0.600000
1963 0.830000 0.800000 1.000000 0.770000
1964 0.920000 1.000000 1.240000 1.000000
1965 1.160000 1.300000 1.450000 1.250000
1966 1.260000 1.380000 1.860000 1.560000
1967 1.530000 1.590000 1.830000 1.860000
1968 1.530000 2.070000 2.340000 2.250000
1969 2.160000 2.430000 2.700000 2.250000
1970 2.790000 3.420000 3.690000 3.600000
1971 3.600000 4.320000 4.320000 4.050000
1972 4.860000 5.040000 5.040000 4.410000
1973 5.580000 5.850000 6.570000 5.310000
1974 6.030000 6.390000 6.930000 5.850000
1975 6.930000 7.740000 7.830000 6.120000
1976 7.740000 8.910000 8.280000 6.840000
1977 9.540000 10.260000 9.540000 8.729999
1978 11.880000 12.060000 12.150000 8.910000
1979 14.040000 12.960000 14.850000 9.990000
1980 16.200000 14.670000 16.020000 11.610000
Scan和read.table不一样。Scan 生成的是有维度的向量，read.table生成的则是带有维度的数据架构.
> time(jj)
Qtr1 Qtr2 Qtr3 Qtr4
1960 1960.00 1960.25 1960.50 1960.75
1961 1961.00 1961.25 1961.50 1961.75
1962 1962.00 1962.25 1962.50 1962.75
1963 1963.00 1963.25 1963.50 1963.75
1964 1964.00 1964.25 1964.50 1964.75
1965 1965.00 1965.25 1965.50 1965.75
1966 1966.00 1966.25 1966.50 1966.75
1967 1967.00 1967.25 1967.50 1967.75
1968 1968.00 1968.25 1968.50 1968.75
1969 1969.00 1969.25 1969.50 1969.75
1970 1970.00 1970.25 1970.50 1970.75
1971 1971.00 1971.25 1971.50 1971.75
1972 1972.00 1972.25 1972.50 1972.75
1973 1973.00 1973.25 1973.50 1973.75
1974 1974.00 1974.25 1974.50 1974.75
1975 1975.00 1975.25 1975.50 1975.75
1976 1976.00 1976.25 1976.50 1976.75
1977 1977.00 1977.25 1977.50 1977.75
1978 1978.00 1978.25 1978.50 1978.75
1979 1979.00 1979.25 1979.50 1979.75
1980 1980.00 1980.25 1980.50 1980.75
> plot(jj, ylab="Earnings per Share", main="J & J")
filter convolution 卷积方法做线性过滤 （相当于移动平均法）
recursive 递归方法做线性过滤 (相当于自回归法AR）
> k = c(.5,1,1,1,.5)
> (k = k/sum(k))
[1] 0.125 0.250 0.250 0.250 0.125
> fjj = filter(jj, sides=2, k)
> fjj
Qtr1 Qtr2 Qtr3 Qtr4
1960 NA NA 0.64500 0.64000
1961 0.65625 0.67875 0.70625 0.73000
1962 0.74000 0.74625 0.76625 0.78375
1963 0.79750 0.82875 0.86125 0.89750
1964 0.95250 1.01125 1.07000 1.13750
1965 1.20125 1.25875 1.30250 1.32500
1966 1.38625 1.47625 1.54875 1.60875
1967 1.63125 1.66500 1.70250 1.76250
1968 1.88625 1.99875 2.12625 2.25000
1969 2.34000 2.38500 2.46375 2.66625
1970 2.91375 3.20625 3.47625 3.69000
1971 3.88125 4.01625 4.23000 4.47750
1972 4.65750 4.79250 4.92750 5.11875
1973 5.41125 5.71500 5.88375 6.00750
1974 6.12000 6.23250 6.41250 6.69375
1975 6.97500 7.12125 7.25625 7.50375
1976 7.70625 7.85250 8.16750 8.56125
1977 8.88750 9.28125 9.81000 10.32750
1978 10.87875 11.22750 11.52000 11.90250
1979 12.35250 12.82500 13.23000 13.71375
1980 14.07375 14.42250 NA NA
lowess 平滑,局部加权多项式回归
> plot(jj) > lines(fjj, col="red") > lines(lowess(jj), col="blue", lty="dashed")
diff 计算差分（差分就是通过做减法得到一个增量的序列）
方差（方差是衡量一个变量波动性的指标）
log 对数
我们把所有jj数据都取log值。
第二步，
我们把log值做差，即使用log值数列中第二值减去第一值，第三值减去第二值，第四值减去第三值等等。
如果做差处理前数列里有n个数值，处理后的结果中将有n-1个数值。
> dljj = diff(log(jj))
> dljj
Qtr1 Qtr2 Qtr3 Qtr4
1960 -0.119545151 0.299516530 -0.658461623
1961 0.326684230 0.123232640 0.287682072 -0.514455392
1962 0.269332934 0.067139303 0.177983155 -0.427444015
1963 0.324496046 -0.036813973 0.223143551 -0.261364764
1964 0.177983155 0.083381609 0.215111380 -0.215111380
1965 0.148420005 0.113944259 0.109199292 -0.148420005
1966 0.007968170 0.090971778 0.298492989 -0.175890666
1967 -0.019418086 0.038466281 0.140581951 0.016260521
1968 -0.195308752 0.302280872 0.122602322 -0.039220713
1969 -0.040821995 0.117783036 0.105360516 -0.182321557
1970 0.215111380 0.203598955 0.075985907 -0.024692613
1971 0.000000000 0.182321557 0.000000000 -0.064538521
1972 0.182321557 0.036367644 0.000000000 -0.133531393
1973 0.235314087 0.047252885 0.116072171 -0.212921997
1974 0.127155175 0.057987258 0.081125545 -0.169418152
1975 0.169418152 0.110541874 0.011560822 -0.246400413
1976 0.234839591 0.140772554 -0.073331273 -0.191055237
1977 0.332705754 0.072759354 -0.072759354 -0.088728230
1978 0.308091059 0.015037877 0.007434978 -0.310154928
1979 0.454736157 -0.080042708 0.136132174 -0.396415273
1980 0.483426650 -0.099206650 0.088033349 -0.321971146
> plot(dljj)
> shapiro.test(dljj)
Shapiro-Wilk normality test
data: dljj
W = 0.9725, p-value = 0.07211
用qqnorm()函数绘制正态概率图 qqline()一条拟合曲线
用hist()函数可以绘制直方图
http://www.stathome.cn/manual/s/10.html
> par(mfrow=c(2,1)) > hist(dljj, prob=TRUE, 12) > lines(density(dljj)) > qqnorm(dljj) > qqline(dljj)
