Reads the footer of a file written by a5_write_pyramid(), its index and
Parquet layout, without reading any rows. This is what a browser fetches
first, so footer_bytes is the cost of opening the file there.
Value
A list of class a5pyramid_info:
path,bytes(file size),footer_bytes(Parquet footer, index included).format_version,generator,a5_crate(the a5 crate that built the file) anda5_index(the A5 index version of its ids).data_resolution: the leaves' resolution.lods: every level's LOD, ascending, the leaves last.tile_size: the most rows in a tile.row_group_size: in a file whose row groups hold several tiles (format version 4), the most rows in a row group;NAwhere each row group is one tile.row_groups: the number of row groups.weights: the weights column (a list of itsnameandtype; format version 4), orNULL.owner: whether the file has anownercolumn, a shard whose rows may be partial (format version 4; seea5_write_pyramid_level()).bbox:c(xmin, ymin, xmax, ymax)of the leaves.columns: a tibble of the value columns'name,type,reducer,codec(seea5_pyramid();NAfor a column that stores its values) andsd, the standard deviation that scaled a"rep_child"column (1 where the leaves' standard deviation is 0 or undefined;NAfor other reducers and older files),dim, the elements per cell of a vector column (NAfor scalar columns), andstorage, a vector column's storage ("fixed_binary"or"list").levels: a tibble of each level'slod,rowsand number oftiles.tiles: a tibble with one row per tile:tile(its 0-based index),row_group(the 0-based index of its row group, as Parquet readers count),lod,rows,bytes(compressed size, what a reader fetches: the row group, or the tile's data pages in a row group of several tiles) andxmin,ymin,xmax,ymax.metadata: the file's other key-value metadata, a named character vector.shard: for a shard of a pyramid set (format version 2, written bya5_write_pyramid_set()), a list of itstier, shardcell(hex) and its shardresolution;NULLfor a single file.
Details
Longitudes follow the a5R stack's convention: a box across the
antimeridian has xmax above 180, as a5R::a5_bbox() reports. The file
itself stores such a box with west > east, as GeoJSON does.
See also
a5_read_pyramid() to read rows.
Examples
library(a5R)
centre <- a5_lonlat_to_cell(-3.19, 55.95, resolution = 10)
cells <- a5_uncompact(a5_spherical_cap(centre, 3000), resolution = 15)
path <- tempfile(fileext = ".parquet")
a5_write_pyramid(data.frame(cell = cells, value = runif(length(cells))), path)
info <- a5_pyramid_info(path)
info
#> ── a5pyramid file file29677dca8a7b.parquet ─────────────────────────────────────
#> Size: 17.1 KB (footer 4.6 KB)
#> A5 index: 1.0
#> Leaves: resolution 15, 1,024 cells
#> LODs: 2-15 (14 levels)
#> Rows: 1,426
#> Tiles: 14 row groups (tile size 4096)
#> Columns: value <float64, mean>
#> Bbox: -3.2194, 55.9355, -3.127, 56.0338
#> Written by: a5pyramid 0.0.0.9000, a5 crate 1.0.0
info$levels
#> # A tibble: 14 × 3
#> lod rows tiles
#> <int> <dbl> <int>
#> 1 2 1 1
#> 2 3 1 1
#> 3 4 1 1
#> 4 5 1 1
#> 5 6 2 1
#> 6 7 2 1
#> 7 8 2 1
#> 8 9 3 1
#> 9 10 4 1
#> 10 11 7 1
#> 11 12 22 1
#> 12 13 76 1
#> 13 14 280 1
#> 14 15 1024 1
