Irregular grid from dataframe to raster in R

I have extracted data at points (lat& lon) at irregular interval. Can a irregular grid be created from the extract dataframe? I do not intend to interpolate the data and keep as it is. I am getting an error trying to create a gridded dataset-

      lon     lat   temp
1  8.5261 50.0223 293.40
2 78.4390 17.2290 295.90
3  8.6350 49.9290 282.88
spg <- df
coordinates(spg) <- ~ lon + lat
gridded(spg) <- TRUE

Error in points2grid(points, tolerance, round) : dimension 1 : coordinate intervals are not constant

Solution

We could round the values

library(sp)
spg[1:2] <- lapply(spg[1:2], function(x) as.numeric(round(x)))
coordinates(spg) <- ~ lon + lat
gridded(spg) <- TRUE

-output

> spg
Object of class SpatialPixelsDataFrame
Object of class SpatialPixels
Grid topology:
    cellcentre.offset cellsize cells.dim
lon                 9       69         2
lat                17       33         2
SpatialPoints:
  lon lat
1   9  50
2  78  17
3   9  50
Coordinate Reference System (CRS) arguments: NA 

Data summary:
      temp      
 Min.   :282.9  
 1st Qu.:288.1  
 Median :293.4  
 Mean   :290.7  
 3rd Qu.:294.6  
 Max.   :295.9

data

df <- structure(list(lon = c(8.5261, 78.439, 8.635), lat = c(50.0223, 
17.229, 49.929), temp = c(293.4, 295.9, 282.88)),
 class = "data.frame", row.names = c("1", 
"2", "3"))