library(dplyr)
std_cats <- c("Layer and Material", "Geometry", "General")
custom_df <- lapply(collection, function(obj) {
props_flat <- unlist(obj$properties, use.names = TRUE)
custom_keys <- names(props_flat)[
!grepl(paste(std_cats, collapse = "|"), names(props_flat))
]
if (length(custom_keys) == 0) return(NULL)
data.frame(objectid = obj$objectid,
prop_name = custom_keys,
prop_value = props_flat[custom_keys],
stringsAsFactors = FALSE)
}) |> do.call(what = rbind)
head(custom_df)
#> objectid prop_name prop_value
#> 1 42 Custom.Zone Zone A
#> 2 42 Custom.Construction.Type RC Frame
#> 3 42 Custom.Floor.Area 428.5
#> 4 43 Custom.Zone Zone B13 Attribute Extraction
Layer names and bounding boxes are just the start. DWG files also carry custom properties and block attributes, zone designations, construction types, sheet numbers, column marks, that make the drawing data genuinely useful for analysis. This chapter extracts those properties from the getData() response and reshapes them into analysis-ready data frames.
This chapter uses collection from Layer Structure Analysis.
13.1 Custom Properties
Custom properties appear under non-standard category names in obj$properties. The standard categories to exclude are "Layer and Material", "Geometry", and "General":
13.2 Block Attributes
Block inserts carry named attributes under obj$properties$Attributes:
block_df <- lapply(collection, function(obj) {
if (is.null(obj$properties$Attributes)) return(NULL)
attrs <- obj$properties$Attributes
data.frame(objectid = obj$objectid,
block_name = obj$name,
attr_name = names(attrs),
attr_value = unlist(attrs),
stringsAsFactors = FALSE)
}) |> do.call(what = rbind)
head(block_df)
#> objectid block_name attr_name attr_value
#> 1 88 TITLE-BLOCK SHEET_NO A-001
#> 2 88 TITLE-BLOCK REVISION Rev 3
#> 3 89 COLUMN-TAG MARK C1
#> 4 89 COLUMN-TAG GRID A/113.3 Pivot to Wide Format
Long-form property data is easy to filter but awkward to join. pivot_wider() gives you one column per property:
library(tidyr)
custom_wide <- custom_df |>
mutate(prop_name = sub("^Custom\\.", "", prop_name)) |>
pivot_wider(names_from = prop_name,
values_from = prop_value)
head(custom_wide)
#> # A tibble: 3 × 4
#> objectid Zone Construction.Type Floor.Area
#> <int> <chr> <chr> <chr>
#> 1 42 Zone A RC Frame 428.5
#> 2 43 Zone B Steel Frame 312.0
#> 3 51 Zone A Masonry 198.313.4 Join Attributes to Layer Data
Merge custom properties back onto the geometry data frame for combined analysis:
enriched <- props_df |>
left_join(custom_wide, by = "objectid")13.5 Floor Area by Zone and Construction Type
zone_summary <- enriched |>
filter(layer == "A-BLDG", !is.na(Zone)) |>
mutate(floor_area = as.numeric(Floor.Area)) |>
group_by(Zone, Construction.Type) |>
summarise(total_floor_area = sum(floor_area, na.rm = TRUE),
n_units = n(),
.groups = "drop")
zone_summary
#> # A tibble: 3 × 4
#> Zone Construction.Type total_floor_area n_units
#> 1 Zone A Masonry 198.3 1
#> 2 Zone A RC Frame 428.5 1
#> 3 Zone B Steel Frame 312.0 113.6 Visualising Custom Attributes
A stacked bar chart of floor area by zone and construction type:
Warning: package 'ggplot2' was built under R version 4.4.3ggplot(zone_summary,
aes(x = Zone, y = total_floor_area, fill = Construction.Type)) +
geom_col(position = "stack") +
geom_text(aes(label = paste0(round(total_floor_area, 0), " m²")),
position = position_stack(vjust = 0.5),
colour = "white", size = 3.5) +
scale_fill_manual(values = c("#367ABF", "#4CAF50", "#F4A261")) +
labs(title = "Floor Area by Zone and Construction Type",
subtitle = "Extracted from block attributes — aerial.dwg",
x = NULL, y = "Floor area (m²)", fill = "Construction type") +
theme_minimal()
13.7 Export
Write the enriched data frame to CSV for handover or further processing:
write.csv(enriched, file = "drawing_attributes.csv", row.names = FALSE)