Updated Comparison Tool Parts 1-3: Refactored for New ALARA Output Parser by eitan-weinstein · Pull Request #154 · svalinn/ALARA

eitan-weinstein · 2025-11-10T19:45:38Z

Closes #168 .
Closes #169 .

This PR implements changes made to tools/alara_output_parser.py for alarajoy_QA, as far as the most up-to-date version, as well as the newest plotting suggestions otherwise in #144 . This PR partially completes #151, however, I will be making adjustments to PRs #145, #146, and #147 as well to get them up to date with the new methods developed in #153.

gonuke · 2025-11-12T21:57:39Z

Does this need a rebase for the output_parser additions?

gonuke

Just reviewed things in the QA script here while we wait for a rebase.

gonuke · 2025-11-13T16:25:30Z

tools/ALARAJOYWrapper/alarajoy_QA.py

+    else:
+        element, A = isotope.split('-')
+        element = element.capitalize()
+        return f'$^{{{A}}}\\mathrm{{{element}}}$'


This might be simpler and equivalent since mathrm is basically the same as non-math

Suggested change

return f'$^{{{A}}}\\mathrm{{{element}}}$'

return f'$^{{{A}}}${element}'

gonuke · 2025-11-13T16:36:16Z

tools/ALARAJOYWrapper/alarajoy_QA.py

+    # Preprocess data to user specifications
+    all_labels = set()
+    all_data = []
+    for df_dict in df_dicts:
+        adf = df_dict[data_key]
+        times = adf.process_time_vals(seconds=seconds)
+        adf = adf.T
+        for col in adf.columns:
+            label_text = f"{adf[col].iloc[0]}"
+            all_labels.add(label_text)
+
+        all_data.append((df_dict, adf, times))
+
+    labels_sorted = sorted(all_labels)
+
+    cmap = plt.cm.get_cmap('Dark2')
+    color_map = {lbl: cmap(i % cmap.N) for i, lbl in enumerate(labels_sorted)}


This might be good in a function of its own

gonuke · 2025-11-13T16:39:48Z

tools/ALARAJOYWrapper/alarajoy_QA.py

+            color = color_map[label_text]
+            label = label_text
+            if data_comp:
+                label = f"{label_text} ({df_dict['Run Label']})"
+
+            ax.plot(
+                times,
+                list(adf[col])[1:],
+                label=label,
+                color=color,
+                linestyle=linestyle,
+            )


Suggested change

color = color_map[label_text]

label = label_text

if data_comp:

label = f"{label_text} ({df_dict['Run Label']})"

ax.plot(

times,

list(adf[col])[1:],

label=label,

color=color,

linestyle=linestyle,

)

label_suffix = ""

if data_comp:

label_suffix = f" ({df_dict['Run Label']})"

ax.plot(

times,

list(adf[col])[1:],

label=label_text + label_suffix,

color=color_map[label_text],

linestyle=linestyle,

)

gonuke · 2025-11-13T16:42:04Z

tools/ALARAJOYWrapper/alarajoy_QA.py

+
+    ax.set_title(title_prefix + title_suffix)
+    if not relative:
+        ax.set_ylabel(f'{df_dict['Variable']} [{df_dict['Unit']}]')


What's the ylabel if it is relative?

gonuke

Thanks for all the work on this. I have a lot of high-level thoughts about the data model, and maybe they'll evolve over the development of this capability.

gonuke · 2025-11-14T16:37:59Z

tools/ALARAJOYWrapper/pyalara.py

+import subprocess
+from string import Template
+from pathlib import Path
+import matplotlib.pyplot as plt


no longer needed

tools/alara_output_processing/alara_output_plotting.py

gonuke · 2025-11-14T16:45:12Z

tools/alara_output_processing/alara_output_plotting.py

+        adf = df_dict[data_key]
+        times = adf.process_time_vals(seconds=seconds)
+        adf = adf.T
+        for col in adf.columns:


Now that this is transposed, each column is a different nuclide, right? (and possibly a total)

Maybe we can note that:

Suggested change

for col in adf.columns:

for nuc in adf.columns:

If we didn't transpose first, could we just perform this operation on all the entries in the first column? The transpose is not a necessary step yet (although it may make life easier for plotting)

gonuke · 2025-11-14T16:51:32Z

tools/alara_output_processing/alara_output_plotting.py

+            (Defaults to True)
+
+    Returns:
+        all_data (list of tuples): A list of all relevant data for each table


What if instead of returning a list of tuples, this method added a new entry for the times to the df_dict for each entry in df_dicts, and collected the labels to make the color_map.

Notwithstanding the discussion about putting some of the metadata into the tables, I like the idea of having the df_dicts cache the processed metadata rather than making new data structures that may be less transparent.

gonuke · 2025-11-14T16:56:16Z

tools/alara_output_processing/alara_output_plotting.py

+    # Plot data
+    for i, (df_dict, adf, times) in enumerate(all_data):
+        linestyle = line_styles[i % len(line_styles)]
+        for col in adf.columns:


These columns represent different nuclides

Suggested change

for col in adf.columns:

for nuc in adf.columns:

gonuke · 2025-11-26T15:03:33Z

It looks like you tried a rebase here, but all the changes from the previous PR are still here.
Maybe its easier to just merge main into this branch.

eitan-weinstein · 2025-11-26T17:11:27Z

It looks like you tried a rebase here, but all the changes from the previous PR are still here. Maybe its easier to just merge main into this branch.

I just rebased again to upstream/main and I think it looks right now with the most recent changes.

gonuke

This looks pretty good.

Mostly minor comments here.

gonuke · 2025-11-26T18:29:11Z

tools/alara_output_processing/alara_output_plotting.py

+    if head:
+        sort_by_time = aop.extract_time_vals([sort_by_time])[0]
+        piv = piv.sort_values(sort_by_time, ascending=False).head(head)


why does this sorting only happen if head?

the sorting would be a good stand alone function as I anticipate a user wanting to sort a dataframe and although it's only a couple of lines, it's not super obvious how

gonuke · 2025-11-26T18:38:37Z

tools/alara_output_processing/alara_output_plotting.py

+        for nuc in piv.index:
+            all_nucs.add(nuc)


Suggested change

for nuc in piv.index:

all_nucs.add(nuc)

all_nucs.add(set(piv.index))

gonuke · 2025-11-26T18:39:35Z

tools/alara_output_processing/alara_output_plotting.py

+            all_nucs.add(nuc)
+
+    nucs_sorted = sorted(all_nucs)
+    cmap = plt.cm.get_cmap('Dark2')


Maybe add a cmap as an input to this function, with default of 'Dark2'?

gonuke · 2025-11-26T18:40:18Z

tools/alara_output_processing/alara_output_plotting.py

+    nucs_sorted = sorted(all_nucs)
+    cmap = plt.cm.get_cmap('Dark2')
+
+    return {lbl: cmap(i % cmap.N) for i, lbl in enumerate(nucs_sorted)}


Might not need nucs_sorted if it's only used once

Suggested change

return {lbl: cmap(i % cmap.N) for i, lbl in enumerate(nucs_sorted)}

return {lbl: cmap(i % cmap.N) for i, lbl in enumerate(sorted(all_nucs))}

gonuke · 2025-11-26T18:42:44Z

tools/alara_output_processing/alara_output_plotting.py

+    else:
+        element, A = isotope.split('-')
+        element = element.capitalize()
+        return f'$^{{{A}}}${element}'


my hero! 🦸‍♂️

gonuke · 2025-11-26T18:45:39Z

tools/alara_output_processing/alara_output_plotting.py

+    for run_lbl, times, filtered, piv, linestyle in data_list:
+


Suggested change

for run_lbl, times, filtered, piv, linestyle in data_list:

for run_lbl, times, filtered, piv, linestyle in data_list:

gonuke · 2025-11-26T18:46:23Z

tools/alara_output_processing/alara_output_plotting.py

+        )
+        data_list.append((run_lbl, times, filtered, piv, linestyle))
+
+    color_map = build_color_map([piv for (_, _, _, piv, _) in data_list])


Suggested change

color_map = build_color_map([piv for (_, _, _, piv, _) in data_list])

color_map = build_color_map([data[3] for data in data_list])

gonuke · 2025-11-26T18:48:36Z

tools/alara_output_processing/alara_output_plotting.py

+                sorted(times),
+                piv.loc[nuc].tolist(),


Sorting the times separately from the values seems fragile

tools/alara_output_processing/alara_output_plotting.py

gonuke · 2025-11-26T18:50:53Z

tools/alara_output_processing/alara_output_plotting.py

+
+    if data_comp:
+        title_prefix = (
+            f'{run_lbls[0]}, {run_lbls[1]} Comparison: \n'


Do you want to support more than 2?

gonuke

LGTM - thanks @eitan-weinstein

gonuke requested changes Nov 13, 2025

View reviewed changes

eitan-weinstein force-pushed the alarajoy_std_plots branch from 8fc31b0 to 6e00665 Compare November 13, 2025 16:54

gonuke requested changes Nov 14, 2025

View reviewed changes

eitan-weinstein force-pushed the alarajoy_std_plots branch 2 times, most recently from 1c1f99b to cd6e1d3 Compare November 19, 2025 20:19

gonuke added this to the ALARA postprocessing tools milestone Nov 21, 2025

eitan-weinstein force-pushed the alarajoy_std_plots branch 4 times, most recently from 1bad73f to 37c67e8 Compare November 25, 2025 23:02

Eitan Shai Weinstein and others added 8 commits November 26, 2025 11:09

Comp tool refactored for new output parser.

8385a7b

Returning missing markdown cell

b71b94c

Fixing df_dict description

4ecb223

Adjusting after rebase.

edde02e

Responding to requested changes.

bd9e12c

Separating out plotting functions, ALARA Python runs, QA

7bf9c5d

Updating for changes made to alara_output_processing

d1aa012

Responding to requested changes.

1867789

eitan-weinstein force-pushed the alarajoy_std_plots branch from 37c67e8 to 1867789 Compare November 26, 2025 17:09

Fixing unit keyword

352fdf5

gonuke reviewed Nov 26, 2025

View reviewed changes

Eitan Shai Weinstein added 2 commits December 1, 2025 10:08

Responding to requested changes.

19a29a3

Fixing time conversion implementation

fb8f586

gonuke approved these changes Dec 1, 2025

View reviewed changes

gonuke merged commit acb98fb into svalinn:main Dec 1, 2025

gonuke mentioned this pull request Dec 1, 2025

Comparison Tool Part 4: Pie Charts #145

Closed

	return f'$^{{{A}}}\\mathrm{{{element}}}$'
	return f'$^{{{A}}}${element}'

	for nuc in piv.index:
	all_nucs.add(nuc)
	all_nucs.add(set(piv.index))

	return {lbl: cmap(i % cmap.N) for i, lbl in enumerate(nucs_sorted)}
	return {lbl: cmap(i % cmap.N) for i, lbl in enumerate(sorted(all_nucs))}

	for run_lbl, times, filtered, piv, linestyle in data_list:


	for run_lbl, times, filtered, piv, linestyle in data_list:

	color_map = build_color_map([piv for (_, _, _, piv, _) in data_list])
	color_map = build_color_map([data[3] for data in data_list])

Conversation

eitan-weinstein commented Nov 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gonuke commented Nov 12, 2025

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

gonuke commented Nov 26, 2025

Uh oh!

eitan-weinstein commented Nov 26, 2025

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

gonuke left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

eitan-weinstein commented Nov 10, 2025 •

edited

Loading