For If loop to add rows of data with copied columns in R

Question

I am trying to recreate an R process that a vendor did for my work. Below is the text "documentation" available:

For any polygon, the data generated by R script needs to be modified before it can be loaded into the database. Follow the following 3 steps to modify the data:

1. Remove all records where Shape="line" (orange fill below)
2. For each of the x,y coordinates where Path ID = 1.1 and Path Order = 1 to n (n=number of sides of polygon), do the following:
   a. Add 2 extra records with Path ID starting from 61 for each record and Path Order = 1 and 2
   b. For 1st record where Path ID = 61 and Path Order = 1, set (0,0) as the (x,y) coordinates. This sets the vertex lines to start from origin
   c. For Path ID = 61 and Path Order = 2, copy the x,y coordinates from Path ID = 1.1 and Path Order = 1. This sets the end-point of the vertex line at the radius greater than 1 d. Repeat this for the other points satisfying criteria defined in step 2. e. Once done for all points, delete all records where Shape = “triangle” and PathID = “1.1”

Img of pre and post rows of data with arrows showing originals in post table

3. The rest of the records for that polygon remain as is.

I have tried to recreate step 2 via a for if loop, but am not able to get the records where 'Path Order'=2 to work:

## prep for loop
newrec61 = NULL
newrec62 = NULL
xcoor62 = NULL
ycoor62 = NULL

## use an for if loop
for (i in 1:10) {
        ## for decagon
if( (dfShapePoints$pathID = 1.1)&(dfShapePoints$pathOrder = i) ) {
    newrec61 <- matrix(ncol=5, nrow=i)
    newrec61[i,]  <- cbind("decagon", 60+i, 1, 0, 0)

   if( (dfShapePoints$pathID = 1.1)&(dfShapePoints$pathOrder = 1) ) {
    ##xcoor62[i,]   <-  dfShapePoints$xCoordinate
}
    if( (dfShapePoints$pathID = 1.1)&(dfShapePoints$pathOrder = 1) ) {
    ##ycoor62[i,]   <-  dfShapePoints$yCoordinate
    }
   ## newrec62 <- matrix(ncol=5, nrow=i)
   ## newrec62[i,]  <- cbind("decagon", 60+i, 2, xcoor62[i], ycoor62[i]))
  ##dfShapePoints <- rbind(dfShapePoints, newrec61, newrec62)
}
}

You can see where I commented out the rest of the work as it is a more complex copy of where I am failing at. I am cbinding 10 entirely null rows and one row of "decagon" "70" "1" "0" "0". Why is my loop not working?

Answer 1

Here is how I would do all the process:

Let's create an initial data frame:

shape = c("square","square","square","square","square","line","line","line","triangle","triangle","triangle","triangle")
PathID = c(1.1,1.1,1.1,1.1,1.1,1,1,1,1.1,1.1,1.1,1.1)
PathOrder = c(1,2,3,4,5,1,2,3,1,2,3,4)
xCoord = c(1,1,2,2,1,5,4,7,2,3,3,2)
yCoord = c(3,4,4,3,3,5,4,7,1,-1,1,1)

df = data.frame(shape,PathID,PathOrder,xCoord,yCoord)

Initial data:

> df
      shape PathID PathOrder xCoord yCoord
1    square    1.1         1      1      3
2    square    1.1         2      1      4
3    square    1.1         3      2      4
4    square    1.1         4      2      3
5    square    1.1         5      1      3
6      line    1.0         1      5      5
7      line    1.0         2      4      4
8      line    1.0         3      7      7
9  triangle    1.1         1      2      1
10 triangle    1.1         2      3     -1
11 triangle    1.1         3      3      1
12 triangle    1.1         4      2      1

Process:

# Keep only PathId == 1.1
df = df[df$PathID == 1.1, ]

# Create a new ID to start at 61 for each polygon
df$NewPathID = c(61, seq(nrow(df)-1) + 61 + cumsum(diff(df$PathOrder)-1))

# Remove the last point for each polygon
df = df[!c(diff(df$NewPathID) < 0, TRUE),]

# Set the new path order to 2
df$NewPathOrder = 2

# Create a new id for each row (will be needed to reorder later)
df$ID = seq(nrow(df))

# create a copy in order to set the path order to 1 and the coordinates to 0
df2 = df
df2$NewPathOrder = 1
df2$xCoord = 0
df2$yCoord = 0

# add df2 as new rows and reorder
df = rbind(df2,df)
df = df[order(df$ID),]

# replace columns and remove temporary ones
df$PathID = df$NewPathID
df$PathOrder = df$NewPathOrder

df$NewPathID = NULL
df$NewPathOrder = NULL
df$ID = NULL

Result:

> df
       shape PathID PathOrder xCoord yCoord
1     square     61         1      0      0
12    square     61         2      1      3
2     square     62         1      0      0
21    square     62         2      1      4
3     square     63         1      0      0
31    square     63         2      2      4
4     square     64         1      0      0
41    square     64         2      2      3
9   triangle     61         1      0      0
91  triangle     61         2      2      1
10  triangle     62         1      0      0
101 triangle     62         2      3     -1
11  triangle     63         1      0      0
111 triangle     63         2      3      1