-
Notifications
You must be signed in to change notification settings - Fork 1
Viewing Output
See Setting Up Input for more information on the HDF5 files used in ACCIV.
When ACCIV is working, it creates a _work directory (the name is set in parameters.ascii) and fills it with intermediate files. When ACCIV completes, it creates two output HDF5 files, so the directory contents look like this after ACCIV finishes a pass:
acciv/tests/syntheticTest/day1/pass1 > ls
outGridVelocity.h5 parameters.ascii
outScatteredVelocity.h5 _work/
The intermediate files can be useful, but we will discuss them later. The two final output files give the velocity field in terms of scattered velocity vectors, and a gridded version based on the scattered vector field. Here is the file structure of the scattered velocity file:
acciv-read-only/tests/syntheticTest/day1/pass1> h5ls outScatteredVelocity.h5
dataX Dataset {49572}
dataY Dataset {49572}
residuals Dataset {49572}
rmsOneSigmaUncertainty Dataset {1}
vx Dataset {49572}
vy Dataset {49572}
x Dataset {49572}
y Dataset {49572}
This file contains six arrays of the same size and one scalar value. The dataX and dataY arrays give the x and y components of the velocity for each velocity vector in units of length and time given by the image bounds and the image time, respectively. The vx and vy arrays are similar, but with the contents of gridGeometryFactors.h5 used to convert from longitude/latitude to meters, if necessary. If the geometry is flat, then dataX and dataY are identical to vx and vy, respectively.
The x and y arrays contain the coordinates of each velocity vector, again in the same units as bounds in the image*.h5 files.
The scattered velocity file has to be plotted to be visualized, since the data are stored just as long one-dimensional arrays. On the other hand, the gridded velocity file is easier to visualize, because the gridded velocity vector maps can be opened as images (right-click in HDFView):
Examples of how to plot the output files in Matlab and Python can be found in:
acciv/tests/syntheticTest> ls plotVel*
plotVelocities.py plotVelocityData.m
