r - Aggregate Columns by weighted average or simple average if weight is missing -

i trying aggregate columns of data frame. in data frame each column correponds industry , each row particular country. idealy aggregate columns weighted average. small fraction weights missing. in case r aggregate country industries simple average. snipet data frame (these exemplary weights other columns)

|   mining   | food     |  weight85| weight90.93| |:----------:|--  -----:|---------:|------------| |   0.9608709| 0.8839236| 0.2738525|   0.1943577| |   0.6445055| 0.8483874| 0.2958678|   0.1043844| |   0.6977353| 0.9449249|        na|          na| |   0.7970192| 0.5941056| 0.2324452|   0.1904089| |   0.7261323| 0.6333187|        na|          na| |   0.9959837| 1.0101725| 0.3872314|   0.1628354| 

i compute weighted average when ingoring missing values problem follows:

  ggpc$mining.weighted <- ggpc$weight85*ggpc$mining   ggpc$food.weighted  <- (1-ggpc$weight85)*ggpc$food   ggpc$food.mining<- rowsums(ggpc[,54:55], na.rm=t) 

building on answer mts provided. came following solution solution, computes 1 row either simple average or weighted average.

 if(sum(is.na(df[1,37])>0)) {1/2*df[1,5]+1/2*df[1,6]}  else  { df[1,37]*ggpc[1,5]+(1-df[1,37])*df[1,6]}  

and further looping through rows of data frame

 df$data.column.agg <- 0  (i in 1:length(df)) {   df[i,*data.column.agg*] <- if(sum(is.na(df[i,*weight column*])>0))  {*simple average* }       else {df[i,*weight column*]*ggpc[i,*data column1*]+(1-df[i,*weight column*])*ggpc[i,*data column2*]}   }