Process Data Viewer (Qt)

Desktop application for interactive analysis of CSV time-series data and WAV signals.

Built with Qt 6 (Widgets + Charts + Concurrent) and powered by a custom C++ analysis library (process-data-toolkit).

This project consists of:

• Process Data Toolkit (core C++ library)
• Process Data Viewer (Qt desktop application)

---

Possible use cases

• analysis of industrial logs
• diagnostics of sensor data
• anomaly detection in control systems

---

Features

CSV Analysis

• Time-series filtering:
  • by sensor
  • by time range (from / to)
• Anomaly detection:
  • Z-score
  • IQR
  • MAD
• Top-N anomaly selection without full recomputation
• Statistics:
  • min / max / mean / stddev
• Data table view with filtering
• Plot with anomaly markers
• Export:
  • JSON report (global or per-sensor)
  • filtered CSV with anomaly markers
  • PNG plot

---

WAV Analysis

• Signal segment selection (windowed analysis)
• Spectrum computation:
  • FFT / DFT / Auto
• Window functions:
  • Hann
  • Hamming
  • None
• Peak detection:
  • local maxima
  • threshold-based
• Top-N peak selection without recomputation
• Statistics:
  • signal min / max / mean / stddev
• Plots:
  • signal waveform
  • frequency spectrum
• Export:
  • PNG plots
  • spectrum CSV
  • text report

---

Demo

CSV Analysis

WAV Analysis

---

UX / Performance

• Asynchronous analysis using QtConcurrent
• UI remains responsive during computation
• Busy state feedback ("Wait for it.")
• Smart recomputation:
  • avoids full recompute when only Top-N changes

---

Underlying Library (Process Data Toolkit)

This application is built on top of a separate C++20 library:

👉 https://github.com/r-lapins/Process-Data-Toolkit

The toolkit provides all core data processing and signal analysis functionality, including:

• CSV parsing and time-series filtering
• Statistical analysis (mean, stddev, quartiles)
• Anomaly detection (Z-score, IQR, MAD)
• WAV signal processing (DFT, FFT, windowing)
• Spectrum computation and peak detection
• CLI tools for batch processing and benchmarking

Why this separation?

The project is intentionally split into:

• PDT (library) → reusable, testable, CLI-capable core
• PDV (this app) → interactive Qt UI on top of the library

This design:

• enforces clean architecture
• enables reuse outside GUI applications
• improves testability and maintainability

---

Architecture

The application is split into three layers:

PDV (Qt UI)
├── Core
│   ├── MainWindow
│   ├── FileLoaderService
│   └── AnalysisTab (factory)
│
├── CSV module
│   ├── CsvAnalysisTab
│   ├── CsvAnalysisControlsWidget
│   ├── CsvAnalysisController
│   ├── CsvAnalysisEngine
│   └── CsvAnalysisResultsPanel
│
├── WAV module
│   ├── WavAnalysisTab
│   ├── WavAnalysisControlsWidget
│   ├── WavAnalysisController
│   ├── WavAnalysisEngine
│   └── WavAnalysisResultsPanel
│
└── PDT (process-data-toolkit)
    ├── csv/
    ├── wav/
    └── core algorithms

Key design decisions

• Controller pattern

  • UI (Tab) does not run analysis directly
  • Controllers handle orchestration + async execution

• Engine layer

  • pure computation (no Qt)
  • easy to test / reuse

• Separation CSV vs WAV

  • similar flow, independent modules

• SessionData

  • unified data entry point
  • contains:
    • CsvData
    • WavData

---

Build Instructions

Requirements

• C++20
• Qt 6 (Widgets, Charts, Concurrent)
• CMake ≥ 3.21

Build

git clone https://github.com/r-lapins/Process-Data-Viewer-Qt/
cd process_data_viewer_qt

git submodule update --init --recursive

mkdir build
cd build

cmake ..
cmake --build .

Run:

./process_data_viewer

---

Usage

Open file

• File → Open
• or Quick Open (predefined examples folder)

CSV workflow

1. Load CSV
2. Adjust:
   • sensor filter
   • time range
   • anomaly method + threshold
3. Toggle:
   • auto update / manual recompute
4. Inspect:
   • table
   • plot
   • alerts
5. Export results

---

WAV workflow

1. Load WAV
2. Select:
   • segment (from, window size)
   • algorithm (FFT / DFT)
   • window function
3. Tune:
   • threshold
   • peak mode
4. Toggle plots:
   • signal
   • spectrum
5. Export:
   • PNG
   • CSV
   • report

---

Example Outputs

• CSV anomaly report (JSON)
• CSV with anomaly markers
• Spectrum CSV (frequency vs magnitude)
• Spectrum report (text)
• Plot PNG exports

---

Notable Implementation Details

• Uses QFutureWatcher + QtConcurrent::run
• Avoids blocking UI thread
• Partial recomputation optimization:
  • tryUpdateTopAnomaliesOnly()
  • tryUpdateDominantPeaksOnly()
• Plot downsampling (performance-safe rendering)
• Clean separation of:
  • UI
  • orchestration
  • computation

---

Project Structure

app/
  ├── core/
  ├── csv/
  ├── wav/
include/pdv/
external/process-data-toolkit/

---

Future Improvements

• Shared base interface for CSV/WAV analysis
• Plugin-style analysis modules
• Unit tests for controllers
• Drag & drop file loading

---

License

MIT (or your chosen license)