The method: a resampling bootstrap | Two modes | Where the weights are stored (DuckDB path) | Identifying rows | Stratified bootstrap weights | Estimating uncertainty | Incremental re-runs | Reuse — nothing to do | More replicates requested | Rows added to the survey table | Forcing a full regeneration | Multiple weight columns | Filtered input tables | Connection lifecycle | Inspecting and removing weights

High-level flow | Stage 1 — Locate or download | Version resolution | Stage 2 — Parse metadata | Format detection | Parsers | SPSS monolithic (parse_spss_mono) | SPSS split-file (parse_spss_split) | SAS reading cards (parse_sas_cards) | LFS codebook CSV (parse_lfs_codebook) | CPSS variables CSV (parse_cpss_csv) | SPSS .sav (parse_spss_sav) | PDF Data Dictionary (parse_pdf_dictionary) | PDF frequency codebook (parse_pdf_codebook) | Metadata encoding | Merge | Stage 3 — Build DuckDB | Data file selection | FWF vs. CSV | Trailing junk row removal (FWF only) | Data fixups (pre-label) | Bootstrap weight join (BSW) | Numeric conversion | Code labels → factors | DuckDB write and ENUM enforcement | Multi-module surveys | LFS pipeline | Connection provenance registry | Registry configuration | Newest-sibling inheritance

Timelines

Naming conventions and where to put the files | Smart defaults and the newest-sibling fallback | When the automatic import fails | See what is actually in the directory | Parse the metadata in isolation | Start from an existing entry as a template | Tweak fixups for data-level issues | Build the full table | Promote the configuration into the registry | Summary

Forced moves

Loading the primary module | Opening a sibling module | Joining modules for analysis | A second example: the Survey of Household Spending | Cleaning up | Database connections | Notes

Loading the primary module | Opening a sibling module | Joining modules for analysis | A second example: the Survey of Household Spending | Cleaning up | Database connections | Notes

The method: a resampling bootstrap | Two modes | Where the weights are stored (DuckDB path) | Identifying rows | Stratified bootstrap weights | Estimating uncertainty | Incremental re-runs | Reuse — nothing to do | More replicates requested | Rows added to the survey table | Forcing a full regeneration | Multiple weight columns | Filtered input tables | Connection lifecycle | Inspecting and removing weights

High-level flow | Stage 1 — Locate or download | Version resolution | Stage 2 — Parse metadata | Format detection | Parsers | SPSS monolithic (parse_spss_mono) | SPSS split-file (parse_spss_split) | SAS reading cards (parse_sas_cards) | LFS codebook CSV (parse_lfs_codebook) | CPSS variables CSV (parse_cpss_csv) | SPSS .sav (parse_spss_sav) | PDF Data Dictionary (parse_pdf_dictionary) | PDF frequency codebook (parse_pdf_codebook) | Metadata encoding | Merge | Stage 3 — Build DuckDB | Data file selection | FWF vs. CSV | Trailing junk row removal (FWF only) | Data fixups (pre-label) | Bootstrap weight join (BSW) | Numeric conversion | Code labels → factors | DuckDB write and ENUM enforcement | Multi-module surveys | LFS pipeline | Connection provenance registry | Registry configuration | Newest-sibling inheritance

Forced moves

Timelines

Naming conventions and where to put the files | Smart defaults and the newest-sibling fallback | When the automatic import fails | See what is actually in the directory | Parse the metadata in isolation | Start from an existing entry as a template | Tweak fixups for data-level issues | Build the full table | Promote the configuration into the registry | Summary

Get a list of property tax related datasets | Get metadata for tax report | Get an overview of land and building values in RS zones | Get data for property tax report and property polygons | Compute and plot relative land values

Background | Example usage: constructing a multi-year series of families in low-income status

Background | Example usage: buildings unoccupied vs not occupied by usual residents

Cancensus and CensusMapper | API Key | Installing cancensus | Accessing Census Data | Census Datasets | Census Regions | Census Geographic Levels | Working with Census Variables | Displaying available Census variables | Variable characteristics | Variable search | Managing variable hierarchy

Census datasets | Variable vectors | View available Census variable vectors | Searching for Census variable vectors | Exact search | Keyword search | Semantic search | Census regions | Standard Geographical Classification | A note on Census Metropolitan Areas and Census Agglomerations | Aside: dissemination areas, blocks, and enumeration areas | Viewing available Census regions | Searching through named Census regions | Exploring Census variable vectors and regions interactively

A simple example | Addendum

Spatial data in cancensus | Maps with base R graphics | Maps with ggplot2 | Interactive maps with leaflet

Word of caution | Ukrainians by Federal Electoral Districts

About | Installing cansim | Usage | License | Attribution

Working with cached tables | parquet | feather | sqlite | Filtering and loading into memory | traditional | Working with the data | Partitioning | Repartitioning | Keeping track of cached data | Removing cached data

Using vectors instead of coordinates

Background | Match between Census Tracts and Census Subdivisions

Population change

Aggregating up data across regions | Change in Vancouver children