2

我想按组计算一堆变量的加权平均值，所以 ddply 和 colwise 似乎是一个合乎逻辑的选择。我的代码是：

library(plyr)

j.parms.by.meas <- ddply(j.subgroups, 
                         .(measure),
                         colwise(weighted.mean, 
                                 .(ee.eer, ee.iplv, sf.per.tstat, kwh.per.sf,
                                   ee.kw.per.sf, ee.kwh.per.sf, efficiency,
                                   motor.eff, ee.lpd, inc.cost), 
                                   j.subgroups$unit.qty))

但是，我收到以下错误：

Error in weighted.mean.default(X[[1L]], ...) : 
  'x' and 'w' must have the same length

看来我将 unit.qty 的整个向量传递给加权平均函数，而不是按照我的意图将 unit.qty 的向量传递给分组变量的子集。有没有一种简单的方法可以将 ddply 生成的子集向量作为参数传递给 ddply 中的指定函数？

以下是一些示例数据 - 请注意，并非所有变量都应为所有度量填充：

                       measure ee.eer ee.iplv sf.per.tstat kwh.per.sf ee.kw.per.sf ee.kwh.per.sf efficiency motor.eff ee.lpd   inc.cost  unit.qty
9813  High Performance Glazing     NA      NA           NA         NA    0.0004788      1.634573         NA        NA     NA   3.177014    817.00
9814  High Performance Glazing     NA      NA           NA         NA    0.0016200      6.031441         NA        NA     NA   0.889816   5051.00
9815  High Performance Glazing     NA      NA           NA         NA    0.0011559      4.607375         NA        NA     NA   1.502042    836.00
9816  High Performance Glazing     NA      NA           NA         NA    0.0006050      2.286861         NA        NA     NA   1.041064   4027.00
9817  High Performance Glazing     NA      NA           NA         NA    0.0008772      3.574747         NA        NA     NA   2.436068 173610.00
9818  High Performance Glazing     NA      NA           NA         NA    0.0010797      4.288593         NA        NA     NA   1.373322   5328.00
9819  High Performance Glazing     NA      NA           NA         NA    0.0010797      4.288593         NA        NA     NA   1.373322   5668.00
9820  High Performance Glazing     NA      NA           NA         NA    0.0010797      4.288593         NA        NA     NA   1.373322    150.00
9821  High Performance Glazing     NA      NA           NA         NA    0.0012480      4.887974         NA        NA     NA   1.190700  33735.40
9822  High Performance Glazing     NA      NA           NA         NA    0.0014988      6.041895         NA        NA     NA   2.081282   3879.50
9823  High Performance Glazing     NA      NA           NA         NA    0.0012100      4.736534         NA        NA     NA   1.126340     69.70
9824  High Performance Glazing     NA      NA           NA         NA    0.0010308      4.307727         NA        NA     NA   2.693508   1356.03
9825  High Performance Glazing     NA      NA           NA         NA    0.0009654      3.810419         NA        NA     NA   1.180242   1832.00
9826  High Performance Glazing     NA      NA           NA         NA    0.0009654      3.810419         NA        NA     NA   1.180242   2748.00
9827  High Performance Glazing     NA      NA           NA         NA    0.0011123      4.275836         NA        NA     NA   0.493198   1722.90
9828  High Performance Glazing     NA      NA           NA         NA    0.0008511      3.332246         NA        NA     NA   0.987162   1850.00
9829  High Performance Glazing     NA      NA           NA         NA    0.0011237      4.651937         NA        NA     NA   2.382166    224.00
9830  High Performance Glazing     NA      NA           NA         NA    0.0010797      4.288593         NA        NA     NA   1.373322    819.00
9831  High Performance Glazing     NA      NA           NA         NA    0.0014532      6.323424         NA        NA     NA   3.401468   3934.00
9832  High Performance Glazing     NA      NA           NA         NA    0.0011559      4.607375         NA        NA     NA   1.502042   9351.00
10329                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    900.00
10330                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768   1246.00
10331                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    265.00
10332                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    308.00
10333                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    500.00
10334                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    335.00
10335                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    550.00
10336                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768    433.00
10337                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    676.00
10338                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    800.00
10339                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768    350.00
10340                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    750.00
10341                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    200.00
10342                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768   1200.00
10343                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    600.00
10344                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768    433.00
10345                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768    500.00
10346                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768    600.00
10347                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9496114     NA 117.631768   4988.00
10348                  >150 HP     NA      NA           NA         NA           NA            NA       0.95 0.9500000     NA 117.631768    250.00

这是来自 dput 的相同数据的转储：

structure(list(measure = structure(c(122L, 122L, 122L, 122L, 
122L, 122L, 122L, 122L, 122L, 122L, 122L, 122L, 122L, 122L, 122L, 
122L, 122L, 122L, 122L, 122L, 124L, 124L, 124L, 124L, 124L, 124L, 
124L, 124L, 124L, 124L, 124L, 124L, 124L, 124L, 124L, 124L, 124L, 
124L, 124L, 124L), .Label = c("(2) 4-lamp 4ft T5HO replacing 1000W MH", 
"(2) 6-lamp 4ft T8 replacing 1000W MH", "2-foot lamp", "2-lamp 4ft T5HO replacing 150W HPS", 
"2-lamp 4ft T5HO replacing 150W MH", "2-lamp 4ft T5HO replacing 175W HPS", 
"2-lamp 4ft T5HO replacing 175W MH", "2-lamp 4ft T8 replacing 150W HPS", 
"2-lamp 4ft T8 replacing 175W HPS", "2-lamp 4ft T8 replacing 175W MH", 
"3-foot lamp", "3-lamp 4ft T5HO replacing 250W HPS", "3-lamp 4ft T5HO replacing 250W MH", 
"4-foot lamp", "4-lamp 4ft T5HO replacing 400W HPS", "4-lamp 4ft T5HO replacing 400W MH", 
"4-lamp 4ft T8 replacing 250W HPS", "4-lamp 4ft T8 replacing 250W MH", 
"6-lamp 4ft T5HO replacing 400W HPS", "6-lamp 4ft T5HO replacing 400W MH", 
"6-lamp 4ft T5HO replacing 750W MH", "6-lamp 4ft T8 replacing 400W HPS", 
"6-lamp 4ft T8 replacing 400W MH", "8-foot lamp", "Automotive Facility", 
"Cold Cathode Fluorecent Lamps 3 w", "Cold Cathode Fluorecent Lamps 5 w", 
"Cold Cathode Fluorecent Lamps 8 w", "Compact flourescent Lamps 13 w Hardwired", 
"Compact flourescent Lamps 14 w Hardwired", "Compact flourescent Lamps 14 w Screw-In", 
"Compact flourescent Lamps 15 w Screw-In", "Compact flourescent Lamps 18 w Hardwired", 
"Compact flourescent Lamps 18 w Screw-In", "Compact flourescent Lamps 20 w Screw-In", 
"Compact flourescent Lamps 23 w Hardwired", "Compact flourescent Lamps 23 w Screw-In", 
"Compact flourescent Lamps 26 w Hardwired", "Compact flourescent Lamps 26 w Screw-In", 
"Compact flourescent Lamps 27 w Hardwired", "Compact flourescent Lamps 27 w Screw-In", 
"Compact flourescent Lamps 32 w Hardwired", "Compact flourescent Lamps 7 w Hardwired", 
"Compact flourescent Lamps 7 w Screw-In", "Compact flourescent Lamps 84 w Hardwired", 
"Court House", "Daylighting Controls", "Dining: Bar Lounge/Leisure", 
"Dining: Cafeteria/Fast Food", "Dining: Family", "Dormitory", 
"Exercise Center", "Exit Signs (Electroluminescent replaceing inc.)", 
"Exit Signs (Electroluminescent replacing CFLs)", "Exit Signs (LED replacing CFL)", 
"Exit Signs (LED replacing incandescent)", "Health Care Clinic", 
"Hospital", "Hotel", "Induction Lighting replacing >=100W and <=200W HPS", 
"Induction Lighting replacing >=70W and <=200W MH", "Induction Lighting replacing >200W and <=250W MH", 
"Induction Lighting replacing >200W and <=400W HPS", "LED Channel Lights", 
"LED traffic lights - Green 12\" - 120V", "LED traffic lights - Green 12\" - 12V", 
"LED traffic lights - Green 8\" - 120V", "LED traffic lights - Green 8\" - 12V", 
"LED traffic lights - Red 12\" - 120V", "LED traffic lights - Red 8\" - 120V", 
"Library", "Manufacturing Facility", "Museum", "Occupancy Sensors", 
"Office", "Parking Garage", "Performing Arts Theater", "Police/Fire Station", 
"Retail (Other Than Mall)", "School/University", "Specialty CFL 14 w Screw-In", 
"Specialty CFL 15 w Screw-In", "Specialty CFL 18 w Screw-In", 
"Specialty CFL 23 w Screw-In", "Specialty CFL 27 w Screw-In", 
"Specialty CFL 7 w Screw-In", "Warehouse", "Workshop", "(2) 4-lamp 4ft T5HO replacing 750W MH", 
"2-lamp 4ft T8 replacing 150W MH", "3-lamp 4ft T8 replacing 250W MH", 
"4-lamp 2ft T5HO replacing 150W MH", "LED Lamps NO Reflector", 
"LED Lamps W/ Reflector", "MR-16 LED Lamps", "Pedestrian W/ countdown", 
"Refrigerated Case LED Lamps NO motion Sensors", "Refrigerated Case LED Lamps W/ motion Sensors", 
"Religious Building", "< 150 Tons", "<= 65 kBtu/h", "> 12 kBtu/h", 
"> 7 and <= 9 kBtu/h", "> 9 and <= 12 kBtu/h", ">= 135 and < 240 kBtu/h", 
">= 150 Tons", ">= 150 Tons and < 300 Tons", ">= 240 kBtu/h", 
">= 300 Tons", ">= 65 and < 135 kBtu/h", "Advanced Diagnostic Tune Up (ADTU) 11-19.9 tons", 
"Advanced Diagnostic Tune Up (ADTU) 20-25 tons", "Advanced Diagnostic Tune Up (ADTU) 4-6 tons", 
"Advanced Diagnostic Tune Up (ADTU) 6-10.9 tons", "Outside Air Economizers", 
"Phase I", "Phase II - Duct Seal", "Programmable Thermostat", 
"Duct Test and Repair (DTR) 11-19.9 tons", "Duct Test and Repair (DTR) 2-6 tons", 
"Duct Test and Repair (DTR) 6-10.9 tons", "High Performance Glazing", 
"Shade Screen", ">150 HP", "1 HP", "1.5 HP", "10 HP", "100 HP", 
"125 HP", "15 HP", "150 HP", "2 HP", "20 HP", "25 HP", "3 HP", 
"30 HP", "40 HP", "5 HP", "50 HP", "60 HP", "7.5 HP", "75 HP", 
"Anti-Sweat Heater Controls", "Beverage Machine Controls", "HiE Compressor - Walk In Cooler", 
"HiE Ice Makers - Air-Cooled - 1001 to 1500lbs", "HiE Ice Makers - Air-Cooled - 101 to 200lbs", 
"HiE Ice Makers - Air-Cooled - 201 to 300lbs", "HiE Ice Makers - Air-Cooled - 301 to 400lbs", 
"HiE Ice Makers - Air-Cooled - 401 to 500lbs", "HiE Ice Makers - Air-Cooled - 501 to 1000lbs", 
"HiE Ice Makers - Water-Cooled - 1001 to 2000lbs", "HiE Ice Makers - Water-Cooled - 101 to 200lbs", 
"HiE Ice Makers - Water-Cooled - 401 to 500lbs", "HiE Ice Makers - Water-Cooled - 501 to 1000lbs", 
"High-Efficiency Evaporator Fan Motors (EC)", "High-Efficiency Evaporator Fan Motors (PSC)", 
"High-Efficiency Freezer (1 Door)", "High-Efficiency Freezer (2 Door)", 
"High-Efficiency Freezer (3 Door)", "High-Efficiency Refrigerator (1 Door)", 
"High-Efficiency Refrigerator (2 Door)", "High-Efficiency Refrigerator (3 Door)", 
"Night Covers", "Snack Machine Controls", "Strip Curtains on Walk-Ins", 
"Computer Power Management - Desktop", "Floating Head Controls", 
"HiE Condenser", "Reach-in Cooler Controls", "CO Sensor / on-off / Office", 
"CO Sensor / VAV / Office", "K-12 School", "Retail", "Dual Technology", 
"Electric or Pneumatic T-Stat", "Existing EMS System", "Integrated Lighting Control", 
"Misc"), class = "factor"), ee.eer = c(NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_), ee.iplv = c(NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_), sf.per.tstat = c(NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_), kwh.per.sf = c(NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_), ee.kw.per.sf = c(0.0004788, 0.00162, 0.0011559, 
0.000605, 0.0008772, 0.0010797, 0.0010797, 0.0010797, 0.001248, 
0.0014988, 0.00121, 0.0010308, 0.0009654, 0.0009654, 0.0011123, 
0.0008511, 0.0011237, 0.0010797, 0.0014532, 0.0011559, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA), ee.kwh.per.sf = c(1.63457252, 6.0314408, 4.60737506, 
2.2868612, 3.57474668, 4.28859278, 4.28859278, 4.28859278, 4.8879738, 
6.04189532, 4.736534, 4.30772732, 3.81041936, 3.81041936, 4.27583642, 
3.33224594, 4.65193678, 4.28859278, 6.32342408, 4.60737506, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA, NA), efficiency = c(NA, NA, NA, NA, NA, NA, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NaN, 0.95, NaN, NaN, 
NaN, NaN, NaN, 0.95, NaN, NaN, 0.95, NaN, NaN, 0.95, NaN, 0.95, 
0.95, NaN, NaN, 0.95), motor.eff = c(NA, NA, NA, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 0.949611437971035, 
0.95, 0.949611437971035, 0.949611437971035, 0.949611437971035, 
0.949611437971035, 0.949611437971035, 0.95, 0.949611437971035, 
0.949611437971035, 0.95, 0.949611437971035, 0.949611437971035, 
0.95, 0.949611437971035, 0.95, 0.95, 0.949611437971035, 0.949611437971035, 
0.95), ee.lpd = c(NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 
NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_
), inc.cost = c(3.177014, 0.889816, 1.502042, 1.041064, 2.436068, 
1.373322, 1.373322, 1.373322, 1.1907, 2.081282, 1.12634, 2.693508, 
1.180242, 1.180242, 0.493198, 0.987162, 2.382166, 1.373322, 3.401468, 
1.502042, 117.631768103245, 117.631768103245, 117.631768103245, 
117.631768103245, 117.631768103245, 117.631768103245, 117.631768103245, 
117.631768103245, 117.631768103245, 117.631768103245, 117.631768103245, 
117.631768103245, 117.631768103245, 117.631768103245, 117.631768103245, 
117.631768103245, 117.631768103245, 117.631768103245, 117.631768103245, 
117.631768103245), unit.qty = c(817, 5051, 836, 4027, 173610, 
5328, 5668, 150, 33735.4, 3879.5, 69.7, 1356.03, 1832, 2748, 
1722.9, 1850, 224, 819, 3934, 9351, 900, 1246, 265, 308, 500, 
335, 550, 433, 676, 800, 350, 750, 200, 1200, 600, 433, 500, 
600, 4988, 250)), .Names = c("measure", "ee.eer", "ee.iplv", 
"sf.per.tstat", "kwh.per.sf", "ee.kw.per.sf", "ee.kwh.per.sf", 
"efficiency", "motor.eff", "ee.lpd", "inc.cost", "unit.qty"), row.names = c(9813L, 
9814L, 9815L, 9816L, 9817L, 9818L, 9819L, 9820L, 9821L, 9822L, 
9823L, 9824L, 9825L, 9826L, 9827L, 9828L, 9829L, 9830L, 9831L, 
9832L, 10329L, 10330L, 10331L, 10332L, 10333L, 10334L, 10335L, 
10336L, 10337L, 10338L, 10339L, 10340L, 10341L, 10342L, 10343L, 
10344L, 10345L, 10346L, 10347L, 10348L), class = "data.frame")

这是我的最终代码，基于 Didzis Elferts 的回答。感谢您的输入！

j.parms.by.meas <- ddply(j.subgroups, 
                         .(measure),
                         function(x) {
                           apply(x[c('ee.eer', 'ee.iplv', 'sf.per.tstat',
                                     'kwh.per.sf', 'ee.kw.per.sf',
                                     'efficiency', 'motor.eff', 'ee.lpd',
                                     'inc.cost')],
                                 2,weighted.mean, x$unit.qty, na.rm=T)
                         })
4

2 回答 2

4

你可以numcolwise在这里使用

ddply(j.subgroups, .(measure),function(x){
  w <- x$unit.qty
  numcolwise(weighted.mean)(x[,-12],w)
})

                  measure ee.eer ee.iplv sf.per.tstat kwh.per.sf ee.kw.per.sf ee.kwh.per.sf efficiency motor.eff ee.lpd   inc.cost
1 High Performance Glazing     NA      NA           NA         NA  0.000977688      3.934811         NA        NA     NA   2.100755
2                  >150 HP     NA      NA           NA         NA           NA            NA        NaN 0.9497194     NA 117.631768
于 2013-02-21T16:54:13.527 回答
3

一种解决方案是colwise()用apply()函数替换。在这种情况下，数据除以变量measure，然后将函数应用于每个子样本。这x只是子样本之一，现在x$unit.qty可以使用。还添加na.rm=TRUE了您的NA值，但这可以删除。

ddply(j.subgroups, 
  .(measure),function(x)
  apply(x[,2:11],2,weighted.mean,x$unit.qty,na.rm=TRUE))

                   measure ee.eer ee.iplv sf.per.tstat kwh.per.sf ee.kw.per.sf ee.kwh.per.sf
1 High Performance Glazing    NaN     NaN          NaN        NaN  0.000977688      3.934811
2                  >150 HP    NaN     NaN          NaN        NaN          NaN           NaN
  efficiency motor.eff ee.lpd   inc.cost
1        NaN       NaN    NaN   2.100755
2       0.95 0.9497194    NaN 117.631768
于 2013-02-21T16:47:48.720 回答