v.sample(1grass)

✨ v.sample — the manual page, but make it cute ✨

okay bestie, welcome to v.sample(1grass) 💖 this is the GRASS User's Manual page for it, and no I will not be normal about it.

🌸 NAME 🌸

v.sample — it samples a raster map at vector point locations. that's it, that's the whole job, and it does it beautifully. 🎀

🔑 KEYWORDS 🔑

vector, sampling, raster — iykyk 🦋

🌈 SYNOPSIS 🌈

v.sample
v.sample --help
v.sample input=name  [layer=string]  column=name output=name raster=name
[method=string]   [zscale=float]   [--overwrite]  [--help]  [--verbose]
[--quiet]  [--ui]

look at that. mostly required args, a couple optional ones, and the classic GRASS flags doing their thing. periodt 💅

🫶 Flags

• --overwrite — lets your output files just bulldoze whatever was already there. existing files? gone. allow output files to overwrite existing files.
• --help — prints the usage summary. your safety net when you forget what anything means. 😭
• --verbose — verbose module output. it will tell you EVERYTHING, and honestly? good for it.
• --quiet — quiet module output. the opposite vibe. shhh. 🤫
• --ui — force launches the GUI dialog. buttons!! ✨

🧚 Parameters

• input=name [required] — name of the input vector point map. or, plot twist, it can be a data source for direct OGR access. main character energy either way. 👑
• layer=string — layer number or name. here's the tea: vector features can have category values in different layers, and this number determines which layer to use. when you're doing direct OGR access, this is instead the layer name. Default: 1
• column=name [required] — name of the attribute column to use for comparison.
• output=name [required] — name for the output vector map where the differences get stored. the receipts, basically. 🍓
• raster=name [required] — name of the raster map to be sampled. the star of the show.
• method=string — sampling interpolation method. the options are: nearest, bilinear, bicubic. Default: nearest
  • nearest — nearest-neighbor interpolation
  • bilinear — bilinear interpolation
  • bicubic — bicubic interpolation
• zscale=float — scaling factor for values read from the raster map. the sampled values get multiplied by this factor. Default: 1.0

---

💜 DESCRIPTION 💜

so v.sample samples a GRASS raster map at the point locations in your input file. and it does it via one of three routes: cubic convolution interpolation, bilinear interpolation, or nearest neighbor sampling (that last one is the default, if you were wondering 💅).

this program can be especially useful when you're sampling for cross validation of interpolations whose output happens to be a raster map. it's giving "the tool you didn't know you needed". 🌟

📝 NOTES 📝

okay here's the important stuff, bestie, don't scroll past:

• the output points will have the easting and northing of the input points. literally the same coords, no cap.
• the input category value is used.
• the input attribute, the raster value, and the difference all get written to output. three-for-one deal. 🔥
• when NULL values are encountered for a cell, a zero value is used instead. in those cases, you might get more acurrate results by using the default nearest neighbor comparisons. (that's spelled the original way, live with it 😌)
• this program may not work properly with lat-long data when the -bc flags are used.
• when interpolation is done (i.e., when the -bc flags are used), values are assumed to be located at the centroid of grid cells. so your current resolution settings matter a LOT. don't ignore them. ⭐

lowkey a warning: lat-long data + -bc flags = possible chaos. and resolution settings when interpolating? not optional to care about. 🙈

🎀 EXAMPLE 🎀

this is a comparison of "elev_ned_30m" and "elev_srtm_30m" North Carolina sample dataset elevation models at random positions. iconic dataset behaviour, honestly. 🦋

# set computational region:
g.region raster=elev_srtm_30m -p
# generate random points:
v.random output=random n=100
# add table with one column:
v.db.addtable random col="elev_srtm30 double precision"
# transfer elevations at random points into table:
v.what.rast map=random rast=elev_srtm_30m col=elev_srtm30
# verify:
v.db.select random
# perform sampling on other elevation map:
v.sample in=random col=elev_srtm30 rast=elev_ned_30m out=elev_samples
#verify:
v.db.select elev_samples
#univariate statistics of differences between elevation maps:
v.univar elev_samples column=diff type=point

see how it flows?? set region, make points, build a table, transfer values, verify, sample, verify again, then univar the diffs. a whole little journey and it ATE. 💅✨

👀 SEE ALSO 👀

• g.region
• v.random
• v.what.rast
• Image Sampling Methods - GRASS Tutorial on s.sample (available as s.sample-tutorial.ps.gz)

💖 AUTHORS 💖

• James Darrell McCauley — when he was at: Agricultural Engineering Purdue University
• Updated for GRASS 5.0 by Eric G. Miller
• Updated for GRASS 5.7 by Radim Blazek

🌈 SOURCE CODE 🌈

available at: v.sample source code (history)

latest change: Friday May 09 14:33:40 2025 in commit: b356c7e108fcbeca77474c8e83ebec8f1e7b3e4a

Main index | Vector index | Topics index | Keywords index | Graphical index | Full index

---

and that's the whole page bestie 🫶 this came out of GRASS 8.5.0 — the GRASS 8.5.0 Reference Manual — with © 2003-2026 GRASS Development Team holding it down. this specific page was last touched Friday May 09 14:33:40 2025 in commit b356c7e108fcbeca77474c8e83ebec8f1e7b3e4a, and it's filed under v.sample(1grass). go sample something. 😭💜