theSA

The Structured Analytics

One desktop app for the whole life of a dataset — clean it, explore it, model it, and visualize it, without ever leaving the same window. Every module keeps working in the background, and any dataset can be sent straight from one module to the next.

Data Cleaning Machine Learning Time Series Visualization Sessions & History Light & Dark Themes

⬇ Download for Windows View Portfolio

What it does

Data Preparation

🧹

Data Cleaning

Fix nulls, remove duplicates, filter and rename columns, handle anomalies and outliers, then export or send the cleaned dataset straight to any other module — no re-uploading.

Modeling

⚠️

Find Anomalies

Detect outliers with Isolation Forest, IQR, Local Outlier Factor, or Mahalanobis distance, with interactive graphs to review every flagged point.

🎯

Predictions

Train classification or regression models, compare every algorithm side-by-side, then predict on new rows one at a time or in bulk from a file.

🎯

Multi Predictions

Train and tune a separate model per output column, compare leaderboards independently, and run bulk predictions across every output at once.

Time Series

🧩

Pattern Detection

Find recurring motifs and unusual regions in a series using matrix-profile analysis — with a live compare view for any two occurrences.

⚡

Sudden Changes

Detect change points and regime shifts with Pelt, Binseg, BottomUp, Window, or Dynp — pick a cost model that matches what "a change" means for your data.

📈

Forecast

Forecast future values with a choice of models, visual confidence intervals, and train/test accuracy comparison before you trust a projection.

Visualization

📊

Plot Graphs

Scatter, line, bar, histogram, box, violin, and more — in X-vs-Y or Trend mode, with reference lines, trend fits, and live zoom/pan.

🔀

One vs All

Compare one main column against every other column at once, with regression lines and correlation badges on every chart.

🗂️

All Columns

Plot every numeric column together — overlayed or in a grid — with normalization so wildly different scales stay comparable.

Specialized

♨

Heat Exchanger Fouling Analysis

Purpose-built analysis for fouling behaviour over time, for both single and multiple heat exchangers.

Workflow

🗂️

Sessions

Save your entire working state — every open module, every loaded dataset — and pick up exactly where you left off, even after closing the app.

📮

Send to Module

Pass a transformed dataset directly into any other module in memory — to one module, several, or all of them — no exporting and re-importing.

🎨

Light & Dark Themes

Switch themes and font sizes to match your setup — every module, table, and chart follows along.

Download theSA