+1
+1
I'm trying to run rasterio in my notebook setup and keep getting when I import the module. Any pointers? I'm running my notebooks inside a container built based on the official jupyter docker stacks. I basically installing the python-numpy gdal-bin libgdal-dev on top of that image.
---------------------------------------------------------------------------
ImportError Traceback (most recent call last)
<ipython-input-1-34708b297916> in <module>()
----> 1 import rasterio
/opt/conda/lib/python3.5/site-packages/rasterio/__init__.py in <module>()
13 import warnings
14
---> 15 from rasterio._base import (
16 eval_window, window_shape, window_index, gdal_version)
17 from rasterio.dtypes import (
ImportError: /usr/lib/libgdal.so.1: undefined symbol: sqlite3_column_table_name
Never mind the above... I had and old version of GDAL. After that, I pulled the --pre version of rasterio and was able to get going.
by some reason i am getting error on from rasterio.vrt import WarpedVRT it shows me an import error.
ImportError: No module named 'rasterio.vrt'
I checked stackoverflow but couldn't find a solution by now. Anyone has the same issue?
Thanks for sharing this wonderful write up. I noticed that AWS_access_key is required, does that mean the landsat data is requester-pay?
Landsat 8 on AWS is not requester-pays.
That is why this works.
http://landsat-pds.s3.amazonaws.com/c1/L8/139/045/LC08_L1TP_139045_20170304_20170316_01_T1/LC08_L1TP_139045_20170304_20170316_01_T1_B5.TIF.ovr
First of all, thanks for the great introduction!
2 Questions:
1.What about GeoTIFFs that are striped instead of tiled? Is it still possible to access a subcrop efficiently?
2. Unfortunately the "with rasterio.Env(CPL_CURL_VERBOSE=True)" changes nothing in the output for me. I'm working on windows with rasterio.version = 1.0.8. Any ideas how I could achieve the verbose output? Been stuck on that for a bit..
Maybe interesting for other windows users: I got an SSL error when trying to access the landsat images. I got rid of it by exposing my SSL certificate with:
import os
os.environ["CURL_CA_BUNDLE"] = r"C:\path\to\curl-ca-bundle.crt"
This is fantastic. I have been attempting to figure out how to manipulate LANDSAT and Sentinel data from AWS without downloading wholesale and without using packages like eo-learn and GEE (too hard to customize for my liking). Had seen these features in rasterio but never knew how they worked. This is a well-done write up, cheers.
jg
Great write up @sgillies - especially cool is the lazy warping