Support scalar-like analytics values by YuanTingHsieh · Pull Request #4683 · NVIDIA/NVFlare

YuanTingHsieh · 2026-05-22T22:16:09Z

What changed

Normalize scalar-like analytics values when constructing AnalyticsData.
Accept Python int/float values as before, plus scalar-like objects with .item() such as numpy.float32 and 0-D numpy arrays.
Normalize numeric values inside SCALARS and METRICS dictionaries, and fail fast if a dict entry is not numeric-scalar-like.
Fix the invalid path validation message to report type(path).
Add unit coverage for generic scalar-like objects, numpy scalar / 0-D array values, METRICS dict normalization, non-scalar dict rejection, and the path error message.

Why

Client tracking writers currently reject values such as numpy.float32 for single scalar metrics because analytics validation only accepts built-in Python float/int. This can happen with code like:

running_loss += cost.cpu().detach().numpy() / images.size()[0]

Centralizing the normalization in AnalyticsData keeps SummaryWriter, MLflowWriter, WandBWriter, and lower-level analytics logging paths consistent without adding a numpy dependency to core API code.

This is a replacement PR for the previous glitched PR #4659, rebased onto the latest NVIDIA/NVFlare:main before branch creation.

Copilot

Pull request overview

This PR updates the AnalyticsData validation layer to accept and normalize “scalar-like” numeric values (e.g., objects with .item() such as NumPy scalars / 0-D arrays), and extends unit tests to cover the new behavior and improved error messaging.

Changes:

Normalize scalar-like values for SCALAR/METRIC analytics values during AnalyticsData construction.
Normalize numeric entries inside SCALARS/METRICS dict payloads and reject non-numeric-scalar dict values.
Add unit tests for scalar-like normalization (including NumPy), dict normalization/rejection, and the path validation error message.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
`nvflare/apis/analytix.py`	Adds numeric-scalar normalization logic and applies it to scalar and dict-based analytics value validation; fixes `path` error message type reporting.
`tests/unit_test/apis/analytix_test.py`	Adds unit coverage for scalar-like and NumPy scalar normalization, dict normalization/rejection, and `path` error message validation.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

greptile-apps · 2026-05-22T22:27:01Z

Greptile Summary

This PR allows analytics tracking to accept scalar-like objects (e.g. numpy.float32, 0-D numpy arrays) in addition to plain Python float/int, normalising them to built-in Python types via a new _normalize_numeric_scalar helper. It also normalises dictionary values for METRICS and SCALARS data types and fixes a pre-existing bug where a falsy-but-not-None path argument (e.g. path=0) bypassed type validation.

_validate_data_types now returns the (possibly normalised) value, which is then stored in self.value, keeping all four writer backends consistent without adding a numpy dependency to the core API.
Dict normalisation is applied only to METRICS/SCALARS (not PARAMETERS), matching the expectation that parameter values can be arbitrary types.
Tests cover generic scalar-like objects, numpy scalars, 0-D arrays, dict normalisation, non-scalar dict rejection, and the corrected path error message.

Confidence Score: 5/5

Safe to merge — the change is additive and well-tested, with no regressions to existing validation paths.

The core normalization logic is self-contained in a new static method, the return-value threading through _validate_data_types is correct, and the path-validation fix is straightforward. All new code paths are exercised by the added tests, including numpy-specific and generic scalar-like cases.

No files require special attention. The _normalize_numeric_scalar static method in analytix.py is the only net-new logic worth a careful read.

Important Files Changed

Filename	Overview
nvflare/apis/analytix.py	Extends _validate_data_types to normalise scalar-like values for SCALAR/METRIC and normalise dict values for METRICS/SCALARS; also fixes a falsy-path guard bug where path=0 previously bypassed validation.
tests/unit_test/apis/analytix_test.py	Adds five new test cases covering scalar-like normalisation, path error message, dict normalisation, non-scalar dict rejection, and numpy scalar/0-D array handling.
tests/unit_test/app_common/widgets/streaming_test.py	Updates the error-message regex in an existing invalid-write test to match the new TypeError text; correct use of raw string for regex escaping.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A["AnalyticsData.__init__(key, value, data_type)"] --> B["_validate_data_types()"]
    B --> C{data_type?}
    C -->|SCALAR or METRIC| D["_normalize_numeric_scalar(value)"]
    D --> D1{float or int?}
    D1 -->|Yes| D2["return value unchanged"]
    D1 -->|No| D3{has callable item method?}
    D3 -->|No| D4["return False - TypeError"]
    D3 -->|Yes| D5{has shape attribute?}
    D5 -->|shape not empty tuple| D4
    D5 -->|shape is empty tuple or absent| D6["call item()"]
    D6 --> D7{result is float or int?}
    D7 -->|Yes| D8["return normalized scalar"]
    D7 -->|No| D4
    D2 --> Z["self.value = normalized value"]
    D8 --> Z
    C -->|METRICS PARAMETERS or SCALARS| E{isinstance dict?}
    E -->|No| E1["TypeError"]
    E -->|Yes| F{METRICS or SCALARS?}
    F -->|No PARAMETERS| G["accept dict as-is"]
    F -->|Yes| H["normalize each dict entry"]
    H --> H1{all entries pass?}
    H1 -->|No| H2["TypeError for offending key"]
    H1 -->|Yes| H3["value = normalized dict"]
    G --> Z
    H3 --> Z
    C -->|TEXT| T{isinstance str?}
    T -->|No| T1["TypeError"]
    T -->|Yes| Z
    C -->|TAGS| U{isinstance dict?}
    U -->|No| U1["TypeError"]
    U -->|Yes| Z

%%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%%
flowchart TD
    A["AnalyticsData.__init__(key, value, data_type)"] --> B["_validate_data_types()"]
    B --> C{data_type?}
    C -->|SCALAR or METRIC| D["_normalize_numeric_scalar(value)"]
    D --> D1{float or int?}
    D1 -->|Yes| D2["return value unchanged"]
    D1 -->|No| D3{has callable item method?}
    D3 -->|No| D4["return False - TypeError"]
    D3 -->|Yes| D5{has shape attribute?}
    D5 -->|shape not empty tuple| D4
    D5 -->|shape is empty tuple or absent| D6["call item()"]
    D6 --> D7{result is float or int?}
    D7 -->|Yes| D8["return normalized scalar"]
    D7 -->|No| D4
    D2 --> Z["self.value = normalized value"]
    D8 --> Z
    C -->|METRICS PARAMETERS or SCALARS| E{isinstance dict?}
    E -->|No| E1["TypeError"]
    E -->|Yes| F{METRICS or SCALARS?}
    F -->|No PARAMETERS| G["accept dict as-is"]
    F -->|Yes| H["normalize each dict entry"]
    H --> H1{all entries pass?}
    H1 -->|No| H2["TypeError for offending key"]
    H1 -->|Yes| H3["value = normalized dict"]
    G --> Z
    H3 --> Z
    C -->|TEXT| T{isinstance str?}
    T -->|No| T1["TypeError"]
    T -->|Yes| Z
    C -->|TAGS| U{isinstance dict?}
    U -->|No| U1["TypeError"]
    U -->|Yes| Z

_{Reviews (5): Last reviewed commit: "Merge branch 'main' into codex/analytics..." | Re-trigger Greptile}

codecov-commenter · 2026-06-30T22:36:38Z

Codecov Report

❌ Patch coverage is 84.61538% with 6 lines in your changes missing coverage. Please review.
✅ Project coverage is 57.01%. Comparing base (4e4ce7e) to head (1830048).

Files with missing lines	Patch %	Lines
nvflare/apis/analytix.py	84.61%	6 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4683      +/-   ##
==========================================
+ Coverage   56.98%   57.01%   +0.02%     
==========================================
  Files         969      969              
  Lines       92310    92344      +34     
==========================================
+ Hits        52606    52650      +44     
+ Misses      39704    39694      -10

Flag	Coverage Δ
unit-tests	`57.01% <84.61%> (+0.02%)`	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

YuanTingHsieh mentioned this pull request May 22, 2026

Support scalar-like analytics values #4659

Closed

YuanTingHsieh changed the title ~~[codex] Support scalar-like analytics values~~ Support scalar-like analytics values May 22, 2026

YuanTingHsieh force-pushed the codex/analytics-numpy-scalars-v2 branch from 7efa912 to b2f76b7 Compare May 22, 2026 22:18

YuanTingHsieh marked this pull request as ready for review May 22, 2026 22:18

Copilot AI review requested due to automatic review settings May 22, 2026 22:18

Copilot started reviewing on behalf of YuanTingHsieh May 22, 2026 22:18 View session

Copilot AI reviewed May 22, 2026

View reviewed changes

Comment thread nvflare/apis/analytix.py Outdated

Comment thread nvflare/apis/analytix.py

Comment thread tests/unit_test/apis/analytix_test.py Outdated

greptile-apps Bot reviewed May 22, 2026

View reviewed changes

Comment thread nvflare/apis/analytix.py Outdated

Comment thread nvflare/apis/analytix.py

YuanTingHsieh force-pushed the codex/analytics-numpy-scalars-v2 branch 2 times, most recently from b4c9ce8 to d66bbbe Compare May 22, 2026 23:57

Support scalar-like analytics values

59f1e2b

YuanTingHsieh force-pushed the codex/analytics-numpy-scalars-v2 branch from 058b94a to 59f1e2b Compare June 30, 2026 22:29

YuanTingHsieh requested a review from holgerroth July 1, 2026 23:00

Merge branch 'main' into codex/analytics-numpy-scalars-v2

1830048

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Support scalar-like analytics values#4683

Support scalar-like analytics values#4683
YuanTingHsieh wants to merge 2 commits into
NVIDIA:mainfrom
YuanTingHsieh:codex/analytics-numpy-scalars-v2

YuanTingHsieh commented May 22, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

greptile-apps Bot commented May 22, 2026 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

codecov-commenter commented Jun 30, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

YuanTingHsieh commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What changed

Why

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

greptile-apps Bot commented May 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Flowchart

Uh oh!

Uh oh!

Uh oh!

codecov-commenter commented Jun 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

YuanTingHsieh commented May 22, 2026 •

edited

Loading

greptile-apps Bot commented May 22, 2026 •

edited

Loading

codecov-commenter commented Jun 30, 2026 •

edited

Loading