Fix time derivative#242
Merged
Merged
Conversation
rouson
force-pushed
the
fix-time-derivative
branch
from
7dbe1e9 to
f0f426e
Compare
This commit updates demo/app/tensor-statistics.F90 to read the model input and output tensor names from training_configuration.json to faciliatate varying the list (e.g., to switch between the simple SB04 microphysics to the production Thompon microphysics). The updates mirror previously-commited changes in demo/app/train-cloud-microphysics.F90.
This update eliminates a name clash caused by other recent versions of the Assert utility.
This commit fixes an issue that was most likely introduced when the bin-filling algorithm was bifurcated into one version below a performance threshold of 80 bins and another version at or above the threshold. The histogram_t bin_frequency_ component was being defined only above the threshold. This commit defines the component outside (after) the performance-threshold conditional branching so the definition happens regardless of which branch executes.
This commit switches the variable name "potential-temperature" to "potential_temperature" in the plot-raw-histograms gnuplot script to match a recent change in the gnuplot file.
This commit finishes the process of adjusting the reading of model input tensors and desired model output tensors so that both are specified by and read from the "tensor names" object in the training_configuration.json file.
This commit 1. Supports constructing an array of file names for tensor inputs and a corresponding array of file names for tensor outputs. 2. Adds a "training data file names" object to the "training_configuration.json" file and a corresponding derived derived type.
With this commit, ./build/run-fpm.sh run tensor-statistics -- --bins 20 will build up an array of training-data file names from information in a "training data file names" JSON object inside the training_configuration.json file. Currently, the tensor-statistics program reads training data only from the first file in the array, computes corresponding input/ouput tensor component histograms, and writes the histograms toa gnuplot (*.plt) files. Known issue: the plots for the input-tensor component histograms appear to be correct, but the output-tensor component histograms are all zeros. To Do ----- 1. Investigate the reason output-tensor values are all zero. 2. Accumulate tensors across the full suite of training-data files. 3. Fix the time-derivative calculation by incorporating the new time_data_t type into the time derivative calculation.
Running demo/app/tensor-statistics.F90 now builds and terminates normally with -DASSERTIONS.
rouson
force-pushed
the
fix-time-derivative
branch
from
f0f426e to
3f83c8b
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
When ready for merging, this pull request will fix an issue identified in the calculation of the desired module outputs, which are time derivatives of a subset of the model inputs, in the demonstration applications (
demo/app). As before, the derivatives are estimated using a first-order-accurate finite difference formula, but the time steps are to be read from a newtime dataJSON object in a new input file that can be written from and read into a newtime_data_tderived type.tensor-statistics.F90demo app to read a set of names of variables from the training data set,dtcalculation intensor-statisticsatime_data_tconstructor (file reader), and