#rioxarray

Want to level up your #Python Dataviz skills? I am launching our Mapping and Data Visualization with Python course on YouTube and designed a free #PythonDatavizChallenge where you spend 30 mins a day for the next 30 days to learn a new skill. Start at

Now that you have learnt how to create globe maps using #CartoPy and Orthographic projection, you can apply the same to the solar eclipse dataset and create a globe visualization for the same.

On Day 10 of #PythonDatavizChallenge, you will learn how to use #rioxarray to load, merge and view geospatial rasters. You will also learn how to add labels to plots using #Matplotlib annotations

Spent the Christmas break playing with the recently released #LISS4 satellite imagery from #ISRO and am happy to share #Python open-source code using #XArray and #Dask for fast parallel processing of large datasets. A thread with details (1/n)

The scenes contain 3 individual GeoTIFFs for each of Green, Red and NIR bands. There's a text file with metadata. The scene swath is 70km and each band is ~500MB of data. (2/n)

Working with large files like these can be tricky with traditional Python libraries like rasterio. You can easily run out of RAM and processing is slow. The #XArray ecosystem makes it a pleasure to work with the data. We read the bands using #rioxarray. (3/n)

How stable are NAIP-derived urban land use classifications overtime? ~Apparently~ not very stable 😓Here we track NAIP pixels that were classified as grass in 2016 and follow their classifications in 2018 and 2020.

Was hoping to see a bit more stability but different timing of imagery and weather may explain that. Note how so much of 2016 grass becomes 2018 trees and 2018 grass becomes 2020 trees. Just putting it out there - thoughts are welcome!

Some reprojection to line the images up. Labels are for the nodes not the paths. #Python #Plotly #Rioxarray @MadreDeZanjas

Blown away how easy the new @OpenDataCube `odc.geo` package by @Kirill_kzb makes it to do normally complicated spatial tasks: for example, plotting large @xarray_dev rasters on an interactive map in only a few lines of code!

The particularly nice thing here is the new standardised `.odc` accessor that's added to every `xarray` object, allowing you to easily inspect, manipulate and transform data loaded by @OpenDataCube and #rioxarray. For example, inspecting geospatial metadata for a dataset:

@William_Lidberg @whiteboxgeo @PyTorch Your workflow is similar to what I do though using a different tool set . I use @pointcloudpipe to convert LiDAR to raster. #rioxarray for selecting subsets of raster and running analytics on them. Parallelized using #Dask . my use case

Happy to announce my new course Mapping and Data Visualization with #Python. This has been in the making for over a year and excited to be able to share it with the world! The full course is free for self-study and shared under an open license. An overview thread below 👇 1/n

We start with the overview of the Python data visualization landscape and zero in on the core libraries for mapping. Day 1 covers vector data visualization with #Maptplotlib, #Pandas, #GeoPandas, and #Contextily (2/n)

Day 2 starts with a deep dive into XArray and raster data visualization. We use #Xarray, #rioxarray and #CartoPy to visualize elevation and gridded climate datasets and learn some advanced #Matplotlib tricks. (3/n)