将R中的每月数据转换为每季度数据

Question

I've created a time series running

debit<-ts(t(data1[7,-1]),start=c(2006,1),freq=12)

then, trying to convert freq=12 to freq=4 using aggregate()

debitq<-aggregate(debit,nfrequency=4)

I get the message

"Error in FUN(newX[, i], ...) : invalid 'type' (character) of argument"

I've tried to include sum as FUN but doesn't work

ts:

           Jan       Feb       Mar       Apr       May       Jun       Jul       Aug       Sep       Oct       Nov       Dec
2006  1151.20   1166.29   1176.28   1138.82   1151.16   1156.34   1154.89   1174.97   1201.12   1204.47   1225.78   1236.95 
2007  1228.42   1260.39   1278.27   1285.94   1299.29   1325.34   1289.39   1312.27   1315.92   1307.14   1329.95   1333.75 
2008  1311.34   1345.38   1356.27   1318.32   1337.18   1343.40   1297.92   1319.52   1335.10   1345.36   1374.38   1397.34 
2009  1352.24   1381.00   1398.24   1384.03   1388.31   1434.60   1457.09   1509.95   1488.93   1472.43   1491.84   1497.39 
2010  1457.66   1494.97   1495.12   1585.07   1614.42   1612.41   1601.38   1618.11   1626.17   1644.93   1666.35   1694.04 
2011  1628.99   1671.78   1695.05   1734.68   1746.29   1805.43   1734.45   1768.39   1808.74   1806.56   1833.54   1866.35 
2012  1801.11   1836.04   1855.93   1880.44   1922.03   1970.70   1876.81   1867.30   1904.96   1943.84   1965.22   2007.98 
2013  1925.81   1951.58   1940.53   1940.31   1935.20   1985.58   1957.00   1991.67   1988.86   2022.52   2069.44   2122.81 
2014  2046.30   2067.28   2080.61   2052.56   2122.92   2202.97   2173.18   2169.54   2183.60   2155.37   2208.96   2295.90 
2015  2247.88   2329.62   2441.20   2451.47   2496.23   2583.69   2603.96   2686.29   2734.63   2646.48   2716.92   2793.01 
2016  2749.88   2819.47   2886.70   2799.79   2878.87   2958.64   2956.51   2955.19   3046.91   3032.89   3092.66   3112.94 
2017  3053.35   3134.67   3234.14   3244.51   3253.03   3357.65   3341.38   3404.00   3430.83   3438.48   3493.38   3559.27 
2018  3528.31   3582.15   3636.33   3658.51   3716.71   3754.10   3748.84   3785.66   3779.48   3763.04   3826.74   3877.06 
2019  3808.26   3873.53   3917.95   3878.69   3890.85   3977.99                                                             

dput():

    V163 = structure(c(1L, 1L, 26L, 1L, 27L, 1L, 15L, 1L, 14L, 
    1L, 4L, 24L, 25L, 23L, 1L, 20L, 13L, 3L, 17L, 1L, 1L, 1L, 
    11L, 10L, 8L, 19L, 7L, 1L, 1L, 9L, 12L, 6L, 1L, 1L, 5L, 18L, 
    22L, 21L, 16L, 2L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
    1L, 1L, 1L, 1L, 1L, 1L), .Label = c("", " 0,00 ", " 1,32 ", 
    " 1.497,49 ", " 1.826,87 ", " 10,06 ", " 10,80 ", " 122,91 ", 
    " 13,59 ", " 138,09 ", " 151,68 ", " 3,53 ", " 3,96 ", " 3.826,31 ", 
    " 3.977,99 ", " 309,25 ", " 36,67 ", " 399,70 ", " 4,38 ", 
    " 410,42 ", " 455,16 ", " 662,76 ", " 84,00 ", " 843,91 ", 
    " 948,54 ", "($)", "Jun/19"), class = "factor")), class = "data.frame", row.names = c(NA, 
-56L))

traceback():

> traceback()
2: apply(array(c(x[1:nend, ]), dim = c(len, nend/len, ncol(x))), 
       MARGIN = c(2L, 3L), FUN = FUN, ...)
1: aggregate.ts(debit, nfrequency = 4)

sapply(debit,class):

> sapply(debit, class)
   7 
"ts" 

Answer 1

From what I can see my best guess is that you load the data using read.csv or read.table from utils . Instead of NA the original file that you were loading into R contained the empty string "" (the first level of the factor in the dput ), which caused the entire column to be seen as character and turned to factors upon loading. In this case you should do someting like this:

data <- read.csv(<YOURFILE>, <YOURARGUMENTS>, colClasses = "character",
 stringsAsFactors = FALSE) # or read.table(... whichever you are using
data[[1]] <- factor(data[[1]])
for(i in 2:ncol(data)) data[[i]] <- as.numeric(data[[i]])