Skip to content

Releases: Kazehaya395/lytrize_desktop

Lytrize Desktop v1.2 ( Windows Release )

Choose a tag to compare

@Kazehaya395 Kazehaya395 released this 07 Sep 17:47
2c77205

Release Notes

Release Highlights

  • 🪟 First official native Windows release
  • 🗺️ Major Map Plot expansion
  • 🌍 World, European, Indian, and US geographic coverage
  • 📍 815 known place names
  • 🖱️ Four new hover annotation controls
  • 🎨 Improved Choropleth color controls
  • ⚡ Improved large-dataset map rendering and sampling
  • 🐛 Multiple Map Plot and Geo Location fixes
  • 📚 Significantly expanded developer documentation

🪟 First Official Windows Release

Lytrize Desktop is now available natively on Windows 10/11 for the first time — alongside the existing Linux (Ubuntu/Fedora) builds.

  • Windows installer: LytrizeSetup_1.2.exe (Inno Setup 7 — installs to Program Files\Lytrize, Start Menu + desktop shortcuts, user data in %APPDATA%\Lytrize)
  • Linux packages: lytrize_1.2_amd64.deb (Debian/Ubuntu) and lytrize-1.2-1.x86_64.rpm (Fedora/RHEL)
  • Full feature parity across both platforms — all charts, Map Plot, Data tools, Dashboards, Sessions & KPIs
  • Offline by design: no internet, no account, no telemetry on either platform
  • Platform-specific install guide covering SmartScreen, UAC, antivirus false positives, plus dual-platform troubleshooting, database reset, and debugging instructions

⚠️ Windows users: SmartScreen may show "Windows protected your PC" on first install — click More info → Run anyway (see README Troubleshooting). Some antivirus tools may flag the unsigned installer as a false positive.


🚀 New Features

Map Plot — Extra Hover Annotations (Choropleth & Scatter)

  • Added a second row in the chart generate/edit panel:
    • Hover Dimension 1
    • Hover Dimension 2
    • Hover Value 1 (+ its own Aggregation)
    • Hover Value 2 (+ its own Aggregation)
  • Dimensions show the most frequent value per location/group.
  • Multiple Values support:
  • Aggregated values are displayed as AGG(col), e.g. MEAN(amount).
  • Works in both:
    • Scatter (Lat/Lon)
    • Choropleth (Location Names)
  • Supported across all three render paths:
    • ISO choropleth
    • World-region fallback
    • City fallback
  • The same column can be used for both Hover Values with different aggregations, e.g.:
    • SUM(amount)
    • MEAN(amount)
  • A column can also be used as both a Dimension and a Value simultaneously.
  • Selections are fully retained by the Edit chart panel.

Map Plot — World Countries, European Regions & India Support

Choropleth Enhancements

  • Added proper ISO-3 country-code support:
    • ISO-2 codes are converted to ISO-3.
    • 3-letter codes are validated against the ISO-3166-1 registry.
    • Invalid codes are rejected instead of rendering blank or incorrect maps.
  • Added 390+ state/province regions across:
    • France
    • Italy
    • Spain
    • Netherlands
    • Switzerland
    • Austria
    • Poland
    • Portugal
    • Belgium
    • Sweden
    • Norway
    • Denmark
    • Finland
    • Czech Republic
    • Hungary
    • Romania
    • Bulgaria
    • Croatia
    • Greece
    • Ireland
    • Slovakia
    • Slovenia
    • Estonia
    • Latvia
    • Lithuania
    • Ukraine
  • Existing regional support retained for:
    • Germany
    • United Kingdom
    • India
    • Canada
    • Australia
  • Added 100+ new cities, including:
    • Lyon
    • Marseille
    • Seville
    • Palermo
    • Innsbruck
    • Salzburg
    • Tampere
    • Aarhus
    • Raipur
    • Kota
    • Jodhpur
    • Amritsar
    • Ludhiana
    • Shimla
    • Dehradun
    • Chandigarh
    • Gurgaon
    • Noida
    • And more
  • Expanded the Geo Location Standardiser remap dropdown from approximately 270 to 815 known place names.
  • Choropleth now auto-detects state, city, and region columns, rather than only country columns.

🐛 Bug Fixes

Map Plot Edit Panel

  • Edit chart selections not retained for new hover dropdowns — fixed.

    • Widget keys were missing from the widget-state specification.
    • Hover selections therefore reset whenever the Edit panel was reopened.
    • All six related selections now persist like the existing chart options.
  • Hover Value 2 silently ignored — fixed.

    • When Hover Value 2 used the same column as Hover Value 1 or a Hover Dimension, it was previously removed by column-name deduplication.
    • Extra hover values now aggregate through temporary column copies, ensuring every requested annotation is rendered.
  • Projection dropdown documentation improved.

    • Now explains that projection is ignored for US-state data, which always uses albers usa.
    • Tile map styles are also documented.

Geo Location Standardiser

  • Misleading success message fixed.
    • "All values resolve cleanly — ready for Map Plot." no longer appears when the column actually failed to resolve.
    • This includes cases where the detected type is:
      • unknown
      • Fewer than 40% matched

Choropleth Location Dropdown

  • Verified state columns missing from dropdown — fixed.
    • The upload optimizer can convert columns such as SC, UT, and NY to category dtype.
    • These columns were previously filtered out of the Location dropdown and therefore invisible to auto-detection.
    • Text detection now supports:
      • object
      • string
      • category
        dtypes.

Choropleth Chart Settings

  • Colorbar and Colorscale controls now work in Choropleth mode.
    • Choropleth figures are now tagged as continuous colour encodings.
    • Styling is applied through Plotly's layout coloraxis, where the choropleth colorbar is actually rendered.
  • Scale option removed from the Generate panel.
    • Scale is now merged with the Colorscale option in Chart Settings.
    • The dropdown is pre-seeded with the scale selected during generation.
  • The following controls now work seamlessly in Choropleth mode:
    • Colorbar show/hide
    • Colorbar title
    • Colorbar tick size
    • Colorbar tick colour
    • Colorbar title size
    • Colorbar title colour
    • Global typography
  • Scatter mode remains unchanged — all existing scatter settings behave exactly as before.

Map Rendering

  • Modern Map API migration

    • Migrated from px.scatter_mapbox() to px.scatter_map() using MapLibre.
    • Added style mapping for:
      • open-street-map
      • carto-darkmatter
      • satellite
    • Fixes blank maps on the bundled Plotly version.
    • All three styles now share the same data-processing pipeline.
  • Categorical colour columns no longer lose points.

    • Removed the location groupby collapse using "first" aggregation.
    • Removed density-cell binning that could overwrite colour columns with point counts.
  • Three-tier sampling

    • >100k points: density aggregation
    • 5k–100k points + categorical colour (≤25 unique values): all points shown
    • 5k–100k points + continuous colour: stratified sampling preserving colour distribution
    • <5k points: all points shown
  • Size-from-Value

    • The Value column now drives marker size when no dedicated Size column is selected, providing Power BI-like behaviour.
  • Honest sampling warnings

    • Added warnings such as:
      • ⚠ X of Y points shown — zoom in for detail
    • Displayed on both tile-map and offline rendering paths.

📚 Documentation

CONTRIBUTOR.md

Expanded into the developer manual for the codebase.

New content includes:

  • Annotated project-structure tree

  • Architecture and data-flow diagram:

    Launcher → Upload → Session → Analysis → Chart runner → Chart settings → Dashboard → HTML export

  • Key subsystem reference covering:

    • Column classification
    • Dtype optimizer
    • Session persistence
    • Transform log & auto-update
    • Chart settings
    • Geo resolution
    • Database
  • Step-by-step "Adding a New Chart Type" guide

  • _WIDGET_SPEC registration requirement

  • Testing section

  • All existing contributor documentation retained

README.md

  • Updated the Map Plot chart-type row to cover:
    • World countries (ISO-3)
    • US states
    • European states/regions
    • Indian states/regions
    • Extra hover annotations

📦 Release Artifacts

Platform Artifact
Windows 10/11 LytrizeSetup_1.2.exe
Debian/Ubuntu lytrize_1.2_amd64.deb
Fedora/RHEL lytrize-1.2-1.x86_64.rpm

Lytrize v1.1

Choose a tag to compare

@Kazehaya395 Kazehaya395 released this 20 Aug 18:49

Lytrize v1.1 — Changelog

🚀 Performance

  • Fixed UI freeze on Full Table preview for large datasets (e.g. 12.7M-row unified Excel tables). The app no longer greys out or stalls when opening Full Table, changing pages, or adjusting filters.
  • Lazy CSV download — the "Download filtered rows" button now only serializes data when actually clicked, instead of generating a multi-hundred-MB CSV on every page rerun.
  • No full DataFrame copies — Full Table pagination now materializes only the visible page (25–500 rows) via a lightweight boolean mask, instead of copying the entire filtered dataset into memory.
  • Faster filter widgets on large files — categorical unique-value scans are sampled (>100k rows) so filter controls render in O(100k) instead of O(rows).
  • O(1) no-filter fast path — opening Full Table with no active filters returns instantly without building a full-row mask.
  • CSV parity — large CSV files benefit from the same optimizations automatically.

📦 Packaging

  • Debian package version bumped to 1.1.
  • RPM package version bumped to 1.1.

🐛 Bug Fixes

  • Resolved the slow-loading / unresponsive upload page that occurred after building a large unified table with joins.

Lytrize-Clip - Chromium Extension

Choose a tag to compare

A Chromium extension that captures full-page screenshots of any webpage including expanded dropdowns, modals, and lazy-loaded content in a single native shot. No scrolling, no zooming, no viewport stitching.

Capture the complete dashboard page directly from Chromium-based browsers and save it as a high-quality image for sharing, reporting, or documentation.

Initial beta release of Lytrize as .deb and .rpm packages.

Choose a tag to compare

Lytrize v1.0 beta is now available as native Linux .deb and .rpm packages.
Some minor unknown bugs might present, please share feedback if found.