Skip to content

Latest commit

 

History

History
130 lines (94 loc) · 6.53 KB

File metadata and controls

130 lines (94 loc) · 6.53 KB

GeoData

Installation Troubles

PYTHONPATH

Minimal

#- export PYTHONPATH=. #- export PYTHONPATH=.:/home/mrilee/git/Fumarole-dabi/fset/build export PYTHONPATH=.:/home/mrilee/git/Fumarole-dabi/fset/build:/home/mrilee/opt/src/mrilee/git/NOGGIN-github/:/home/mrilee/opt/src/mrilee/git/CCL/

export PYTHONPATH=.:/home/mrilee/git/GeoData/geodata/:/home/mrilee/git/GeoData/

export PYTHONPATH=.:/home/mrilee/git/GeoData-dabi/geodata/:/home/mrilee/git/GeoData-dabi/

Old PYTHONPATH

.:/home/mrilee/opt/src/mrilee/git/Fumarole/build-fset:/home/mrilee/opt/src/mrilee/git/STARE/build1/src/:/home/mrilee/opt/src/mrilee/git/NOGGIN-github/:/home/mrilee/opt/src/mrilee/git/CCL/:/home/mrilee/opt/src/mrilee/git/GeoData/

.: home/mrilee/opt/src/mrilee/git/Fumarole/build-fset: /home/mrilee/opt/src/mrilee/git/STARE/build1/src: home/mrilee/opt/src/mrilee/git/NOGGIN-github:/home/mrilee/opt/src/mrilee/git/CCL/: home/mrilee/opt/src/mrilee/git/GeoData

fset compilation

FERMAT_SDK_HOME=/home/mrilee/opt/src/mrilee/Fermat/FermatSDK20200603/

/home/mrilee/opt/src/mrilee/Fermat/FermatSDK20200603-1

PYTHON_INCLUDE_DIRS

export PYTHON_INCLUDE_DIRS=/home/mrilee/opt/anaconda3/envs/2022-work/include/python3.10:/home/mrilee/opt/anaconda3/envs/2022-work/lib/python3.10/site-packages/numpy/core/include/

home/mrilee/opt/anaconda3/envs/2022-work/include/python3.10 /home/mrilee/opt/anaconda3/envs/2022-work/lib/python3.10/site-packages/numpy/core/include

OLD PYTHON_INCLUDE_DIR

PYTHON_INCLUDE_DIRS=/home/mrilee/opt/anaconda3/include/python3.7m/:/home/mrilee/opt/anaconda3/lib/python3.7/site-packages/numpy/core/include/

PYTHON_INCLUDE_DIRS= home/mrilee/opt/anaconda3/include/python3.7m: home/mrilee/opt/anaconda3/lib/python3.7/site-packages/numpy/core/include

Notes on sketches

(insert (concat “\n” (shell-command-to-string “ls geodata/* | grep -v \~ “)))

geodata/geodata.py The main module of support routines.

geodata/sketch1.py Load GOES IMG Band 5 data (NetCDF4)

geodata/sketch10.py Load GOES SNDR Band 7 data

geodata/sketch11.py Load MERRA-2 TPW data, scatterplot

geodata/sketch12.py Cross comparison MERRA-2 and GOES IMG Band 5, plot correlation

geodata/sketch13.py Cross comparison MERRA-2 and GOES IMG Band 3

geodata/sketch14.py Cross comparison MERRA-2 and GOES IMG Band 4

geodata/sketch2.py Synthetic data containing parameterized tracks

geodata/sketch3.py Load and scatterplot MERRA-2 data

geodata/sketch4.py Load and scatterplot AAOP/DMOP/MHOP alternative TPW data (MSPPS ORB)

geodata/sketch5.py Load MERRA-2 times and create STARE temporal indices

A big question here is are we working with intervals or points.

geodata/sketch6.py Work with STARE temporal ids from NetCDF4 datasets

Do we want to work with intervals or points in time?

geodata/sketch7.py Look for intersections between MERRA and GOES files (reads files)

geodata/sketch8.py Get STARE temporal id from filenames and compare (uses filenames)

geodata/sketch9.py Construct a compliant h5 data file.

geodata/sketch90.py Read from the h5 file and check the results of sketch9.py.

geodata/sketch91.py Load GOES dataset from NetCDF4 files, load a few locations and check resolution.

geodata/sketch92.py Load data from the h5 file and fit three components to the histogram.

geodata/sketch93.py Load h5 data, like 92, but display the ratio b5[ge]/b5[lt]

geodata/sketch94.py Load h5 data, like 92, but display image masked by m2 tpw at various thresholds

geodata/sketch94a.py Load h5 data, like 92, but display image masked by b5 with various thresholds

geodata/sketch95.py Load h5 data and try to compare b4-b5 against TPW

Added root fitting to theoretical model that matches linear fit very well. Threshold of about -1850 found.

geodata/sketchA.py Check terminators at various resolutions, esp. 27.

geodata/sketchB.py Go over sortedcontainers and iterators.

geodata/sketchC.py DASK accelerates ps.from_latlon STARE calc.

geodata/sketchD.py Stopwatch sketching,

Don’t forget nx,ny <-> row,column, i.e. ny,nx…

geodata/sketchE.py Try CCL. Used both cv2 calls and ccl_marker_stack.

geodata/sketchE0.py Try to understand cv2.threshold

geodata/sketchE1.py Try CCL and tracking with multiple h5 files. Fixed the color mapping.

geodata/sketchE2.py Some clean up or refactoring of E1. CCL on b4-b5.

geodata/sketchE3.py Refactor CCL/Tracking as a class. Adds various h5 outputs including label resolution.

geodata/sketchF.py Try to make preprocessing into a callable function.

geodata/sketchF0.py Add visualization to check previous.

geodata/sketchF1.py Multi-file load and join, save to h5 files.

geodata/sketchF2.py Load h5 files and display to check.

geodata/sketchG.py Load MODIS data into a dataset

geodata/sketchG0.py Check the MODIS data in h5 format, including GRING

geodata/sketchG1.py Integrate GOES and MODIS data, show comparison viz.

geodata/sketchG1-1.py Show a whole MODIS granule, superimposed on GOES in a lat-lon bbox.

geodata/sketchG2.py Compare GOES and MODIS Granules by “clearing” bits to level, uses GRING, and distribute across nodes

geodata/sketchH0.py Partition MODIS data, basic sketches for VDS

geodata/sketchH1.py Shape and type check of VDS-style data stored in HDF

geodata/sketchH2.py Visualize VDS-style partitioning of MODIS data

geodata/sketchI0.py Intersection example

geodata/sketchJ0.py Implement a new src_coord convention.

geodata/sketchJ1.py Implement a new src_coord convention. Added to GeoData/geodata.

geodata/sketchJ2.py Virtual dataset experiments

geodata/sketchJ3.py Virtual dataset experiments

geodata/sketchK0.py Load GOES & MODIS, lexsort, nadir vs. wing resolution - shows box on lower left and nadir

geodata/sketchK0a.py Load GOES & MODIS, lexsort, nadir vs. wing resolution - plot 2 scans, allow zoom to show overlap

geodata/sketchK1.py GOES+MODIS Template

geodata/sketchK2.py GOES+MODIS - plot 2 scans, 2 subplots, full granule + box on left

geodata/sketchL0.py Rework native array index encoding (src_coord, early version of sketchJ*.py?)

geodata/sketchL1.py Rework native array index encoding (src_coord, early version of sketchJ*.py?)

geodata/sketchM0.py Visualize root polyhedron

Details on performing CCL on the GOES bands

Choose a threshold

Eliminate the sky

OpenCV.connectedComponents works on uint8 (only?)

compare 1 f-dataset to per channel f-dataset