Comments (4)
dionhaefner
commented on July 19, 2024
We should keep an eye on this PR: cogeotiff/rio-cogeo#22
Clipping rasters to the Mercator grid at some maximum zoom level should be the right way to tackle this.
from terracotta.
j08lue
commented on July 19, 2024
Why, when do we currently reproject? When serving?
from terracotta.
dionhaefner
commented on July 19, 2024
Yes. We construct a WarpedVRT when reading the raster data, and we should keep doing that. It's just that there's no real overhead to reproject during ingestion should the user wish to do so, and squeeze out some drops of performance. We just need to be careful not to destroy data this way (e.g. at extreme latitudes).
from terracotta.
dionhaefner
commented on July 19, 2024
Some numbers to back this up:
------------------------------ benchmark 'test_bench_singleband': 24 tests -------------------------------
Name (time in ms) Min Max Median
----------------------------------------------------------------------------------------------------------
test_bench_singleband[True-subpixel-nearest] 10.4022 (1.0) 22.1370 (1.41) 10.9022 (1.0)
test_bench_singleband[True-subpixel-linear] 10.5489 (1.01) 18.4306 (1.17) 10.9121 (1.00)
test_bench_singleband[True-subpixel-average] 10.6505 (1.02) 17.9965 (1.14) 11.0098 (1.01)
test_bench_singleband[True-subpixel-cubic] 10.6557 (1.02) 15.7493 (1.0) 11.0352 (1.01)
test_bench_singleband[True-preview-nearest] 10.7739 (1.04) 18.4937 (1.17) 13.3235 (1.22)
test_bench_singleband[True-preview-average] 11.2968 (1.09) 18.8943 (1.20) 11.6789 (1.07)
test_bench_singleband[True-preview-linear] 11.7683 (1.13) 19.3742 (1.23) 12.1181 (1.11)
test_bench_singleband[True-preview-cubic] 12.0574 (1.16) 18.6076 (1.18) 12.3896 (1.14)
test_bench_singleband[False-preview-nearest] 12.2605 (1.18) 17.6876 (1.12) 12.7577 (1.17)
test_bench_singleband[False-subpixel-nearest] 12.4894 (1.20) 20.1799 (1.28) 12.8388 (1.18)
test_bench_singleband[False-subpixel-linear] 15.9545 (1.53) 32.8731 (2.09) 16.6443 (1.53)
test_bench_singleband[False-subpixel-average] 16.1189 (1.55) 42.0965 (2.67) 16.6298 (1.53)
test_bench_singleband[False-preview-average] 16.4866 (1.58) 23.3200 (1.48) 17.3809 (1.59)
test_bench_singleband[False-preview-linear] 16.6721 (1.60) 23.4794 (1.49) 17.1587 (1.57)
test_bench_singleband[False-subpixel-cubic] 19.0431 (1.83) 26.7898 (1.70) 19.5505 (1.79)
test_bench_singleband[False-preview-cubic] 20.0107 (1.92) 45.5268 (2.89) 20.8642 (1.91)
test_bench_singleband[True-birds-eye-nearest] 20.0978 (1.93) 27.7186 (1.76) 20.7752 (1.91)
test_bench_singleband[True-birds-eye-linear] 20.5290 (1.97) 27.8032 (1.77) 21.2586 (1.95)
test_bench_singleband[True-birds-eye-average] 20.6013 (1.98) 39.2488 (2.49) 22.4438 (2.06)
test_bench_singleband[True-birds-eye-cubic] 21.1571 (2.03) 33.8629 (2.15) 23.9085 (2.19)
test_bench_singleband[False-birds-eye-nearest] 29.3875 (2.83) 35.4783 (2.25) 29.9778 (2.75)
test_bench_singleband[False-birds-eye-average] 53.7812 (5.17) 60.7123 (3.85) 55.1779 (5.06)
test_bench_singleband[False-birds-eye-linear] 54.4519 (5.23) 97.4871 (6.19) 59.9822 (5.50)
test_bench_singleband[False-birds-eye-cubic] 60.2341 (5.79) 66.7784 (4.24) 61.7923 (5.67)
----------------------------------------------------------------------------------------------------------
The cases with True in the first parameter is where it's already in the target CRS. Especially the pathological birds-eye case gets defused pretty well. Not a lot of influence though when using nearest neighbor interpolation.
from terracotta.
Related Issues (20)
- Support multi-band GeoTiffs HOT 6
- New release HOT 8
- Add check whether Terracotta is running in multithreaded context HOT 5
- different table schema in sqlite driver between conda and pip terracotta version HOT 16
- Question: have key wildcard to serve multiple geotiffs as mosaic HOT 1
- Installs without postgres driver? HOT 1
- Ensure that ingest is using sqlite HOT 6
- Postgres driver ignores supplied database name HOT 2
- Postgres driver.create fails when database exists HOT 5
- rasterio error when deploying to AWS lambda with zappa HOT 5
- Question: speed of compute metadata HOT 2
- allowing CORS in AWS Lambda HOT 9
- Error on get_datasets with Postgresql database HOT 2
- getting png from singleband error: Unsupported operand type(s) for /: 'decimal.Decimal' and 'float' HOT 6
- NoSuchTableError: key_names on zappa deploy HOT 2
- Support for non Web Mercator (EPSG 3857) tiles. HOT 3
- Windows 10? HOT 1
- Compatibility with Numpy 1.24 HOT 2
- Current version is not working. HOT 4
- Misleading explicit_color_map documentation HOT 6
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from terracotta.