Migrating from 1.x to 2.0 • metrosp

metrosp 2.0 renames most datasets, several columns, and the cache helpers. The tables below map each 1.x name to its 2.0 name. See the changelog for the reasons behind each change.

Datasets

1.x	2.0
`passengers_entrance`	`line_entries_monthly`
`passengers_transported`	`line_transported_monthly`
`station_averages`	`station_transported_monthly`
`station_daily`	`station_entries_daily`
`lines`	`rail_lines`
`stations`	`rail_stations`
`station_inauguration`	Removed

calendar_spo and metro_colors keep their names.

Columns

1.x	2.0	Datasets
`metric_abb`	`metric`	Demand datasets
`metric`	`metric_name`	Demand datasets
`metric_pt`	`metric_name_pt`	Demand datasets
`passengers`, `avg_passenger`	`value`	Demand datasets
`is_ponto_facultativo`	`is_optional_holiday`	`calendar_spo`
`is_feriadao`	`is_long_weekend`	`calendar_spo`

The station datasets and rail_stations now have station_id, a stable, opaque identifier shared by members of the same physical station complex. Keep station_name for display. Grouping station_transported_monthly by station_id adds boardings across its platforms, not unique entrants. station_transported_monthly also has metric = "mdu", metric_name, and metric_name_pt; station_entries_daily has no metric columns. The year and line_number columns are integers in the demand datasets.

metric changed meaning: it held the English label in 1.x and holds the short code (total, mdu, …) in 2.0. Code that filters on metric needs the new codes.

Functions

1.x	2.0
`metrosp_cache_dir()`	`metrosp_cache()`
`metrosp_cache_list()`	`metrosp_cache()`
`metrosp_cache_enable()`	Removed; caching is on by default

read_metro_demand() and metrosp_cache_clear() keep their names. metrosp_cache() lists cached files and prints the cache location; it does not return only a directory path. Use attr(metrosp_cache(), "directory") to get that path. Set cache = FALSE on a read to use session-temporary storage. Version 2.1 also adds read_metro_service() for line_service_monthly.

Values

line_transported_monthly$value now counts passengers, not thousands. Drop any * 1000 your code applied.
The line_number == 99 network rows are gone, with no replacement.

Example

# 1.x
passengers_entrance |>
  dplyr::filter(metric_abb == "total") |>
  dplyr::select(date, line_number, passengers)

# 2.0
line_entries_monthly |>
  dplyr::filter(metric == "total") |>
  dplyr::select(date, line_number, value)