Creating xts time series

问题

zz <- textConnection("strtimestamp, jstimestamp, 61757, 61754, 61760, 61753, 61758, 61762, 61756, 61759, 61761, 61755, 61752
'01/01/2007 00:00:00', 1167606000000, 1145.3, 1647.3, 612, 963.3, 1063.5, 2294.3, 726.9, 280.8, 5182.4, 739.5, 1631.7
'01/01/2007 00:05:00', 1167606300000, 1116.7, 1625.4, 695.3, 948.4, 1111.8, 2262, 736.5, 202.7, 5135.8, 733, 1608.1
'01/01/2007 00:10:00', 1167606600000, 1100.3, 1613.4, 614.9, 957.5, 1059.8, 2261.6, 723.2, 238.5, 5114.5, 733, 1599.1
'01/01/2007 00:13:41', 1167606821000, 1101.2, 1619.6, 600.9, 947.7, 1047.3, 2254.1, 715.5, 269.7, 5124.3, 726.5, 1608.2
'01/01/2007 00:15:41', 1167606941000, 1097.6, 1607.6, 626.2, 944.2, 1032.2, 2224.9, 715.5, 260.6, 5094.3, 728.9, 1602.3
'01/01/2007 00:25:00', 1167607500000, 1081.1, 1622, 616.7, 955.1, 1047, 2240.1, 717.7, 259.8, 5088.6, 737.8, 1578.7
'01/01/2007 00:30:00', 1167607800000, 1079.5, 1602.3, 601.6, 953.2, 1019.7, 2195.7, 710.8, 266, 5043.4, 742.8, 1560.9
'01/01/2007 00:35:00', 1167608100000, 1055.6, 1597.3, 599.4, 947.9, 1050.8, 2168, 723.9, 242.1, 5063.7, 732.9, 1554.5
'01/01/2007 00:40:00', 1167608400000, 1050.5, 1585.7, 618.4, 942.3, 1025.1, 2142.6, 705.5, 236.6, 5012.4, 745, 1555.3
'01/01/2007 00:45:00', 1167608700000, 1054.4, 1588.8, 607.7, 917.4, 1030.4, 2143.6, 696.9, 246.4, 5011.9, 736.5, 1565.8
'01/01/2007 00:50:00', 1167609000000, 1063, 1580.7, 587.7, 937.5, 1002.2, 2136.6, 694.5, 247.4, 5007.5, 740.8, 1535
'01/01/2007 00:55:00', 1167609300000, 1054.2, 1576.5, 623.6, 932.1, 1017, 2101, 708.6, 208.9, 4999.9, 740.5, 1532.4
'01/01/2007 01:00:00', 1167609600000, 1061.8, 1561.2, 574.1, 906.4, 1011.5, 2102.9, 705.1, 235.7, 4963.5, 729.1, 1546.6
'01/01/2007 01:05:00', 1167609900000, 1066.2, 1563.7, 642.4, 923, 1024.1, 2075.3, 682.5, 200.8, 4945, 738, 1523.2
'01/01/2007 01:10:00', 1167610200000, 1025.4, 1556.3, 577, 931.9, 1009.8, 2064.6, 684.7, 237.2, 4917.2, 725.1, 1503.6
'01/01/2007 01:15:00', 1167610500000, 1022.7, 1545.5, 573, 918, 992.6, 2052.3, 695.2, 224.7, 4926.7, 725.2, 1496
'01/01/2007 01:20:00', 1167610800000, 1018.4, 1544.9, 582.7, 897.7, 985.3, 2039.5, 681.9, 225.2, 4891.7, 728.2, 1534.1
'01/01/2007 01:25:00', 1167611100000, 1012.4, 1529.3, 558.9, 915.4, 969.7, 2029.1, 694, 241.4, 4874.9, 731.9, 1491.4
'01/01/2007 01:30:00', 1167611400000, 1013.1, 1525, 581.9, 911.6, 970.4, 2017.1, 685.2, 231.2, 4866.2, 735.8, 1489.1
'01/01/2007 01:35:00', 1167611700000, 1011.4, 1530.9, 567.4, 890.1, 964.3, 2001, 691.8, 227.8, 4821.6, 735.6, 1492.5
'01/01/2007 01:40:00', 1167612000000, 1006.6, 1523.6, 574.5, 882, 963.9, 1966.6, 662, 228, 4800, 749.3, 1501
'01/01/2007 01:42:55', 1167612175000, 1004, 1514.5, 544.7, 896, 949.5, 1974.1, 682.5, 240, 4814, 735.4, 1480.4
'01/01/2007 01:44:29', 1167612269000, 1007.1, 1513.4, 561.7, 873, 948, 1976.1, 677, 229.4, 4778.9, 723.8, 1492.4
'01/01/2007 01:50:00', 1167612600000, 1015, 1506, 547.7, 895.7, 923.6, 1980.8, 667.7, 238.4, 4772.8, 731.1, 1476.9
'01/01/2007 01:55:00', 1167612900000, 990.2, 1490.8, 582.7, 869.9, 933.7, 1941.4, 663.7, 225.2, 4745.7, 738.5, 1468.8
'01/01/2007 02:00:00', 1167613200000, 978, 1484.9, 569.6, 890.9, 945.7, 1905.7, 679.3, 215.7, 4711.1, 723.8, 1455.3
'01/01/2007 02:05:00', 1167613500000, 994.7, 1489.5, 544.9, 861.1, 934.3, 1927.1, 679.4, 212.7, 4697.3, 717.8, 1468.9
'01/01/2007 02:10:00', 1167613800000, 990.6, 1471.4, 569, 883, 920, 1894.5, 661.4, 215.6, 4672.8, 731.3, 1439.6
'01/01/2007 02:15:00', 1167614100000, 976.5, 1480.8, 531.4, 880.3, 935.9, 1903.4, 659.9, 224.9, 4691.2, 720.1, 1442.9
")
Data <- read.table(zz,  header = TRUE,  sep = ",",  dec=".")

How can I read this time series into xts, zoo or some other R time series class that supports irregular time series?

format of the first column is "%m/%d/%Y %H:%M:%S"

==================

Vincent's first example works so I am accepting his answer. The other two do not work on the real data. It says that index is not unique, even though in my data I do not have duplicate datetime.

回答1:

You can read the data into a data.frame, manually parse the date, and convert it to an xts object. If something goes wrong, you will see more clearly what happens.

d <- read.csv(textConnection(s))
d <- xts( 
  d[,-1], 
  as.POSIXct( 
    as.character(d[,1]), 
    format="'%m/%d/%Y %H:%M:%S'"  # Note the single quotes...
  ) 
)

You could also use read.zoo to parse the data.frame.

d <- read.csv(textConnection(s))
d <- read.zoo(d, 
  format="'%m/%d/%Y %H:%M:%S'", 
  FUN=as.POSIXct  # The default is to use dates, not times
)

You could also use read.zoo directly.

d <- read.zoo(textConnection(s), 
  sep=",", header=TRUE, 
  format="%m/%d/%Y %H:%M:%S",  # No single quotes: read.zoo calls 
  FUN=as.POSIXct               # read.table, not read.csv:   
)                              # it has different default parameters

来源：https://stackoverflow.com/questions/9377424/creating-xts-time-series

标签

time-series

xts

zoo