Data table
Data 1
Up to 20,000 observations per table. Project files open up to 256 MiB. Browser space is shared with other work; keep a project file backup.
Blank responses are missing values. Excluded values have blue italic text. Use Observation details to change IDs, units, SD and exclusions.
Import or paste a table
Import a Column table with one or more response columns (such as Control, Low dose and High dose), or a long-form table with one row per observation. Choose the layout after reading the file. Include group, response, dose, unit, and replicate ID as needed. Blank responses stay missing. Use a point or comma for a decimal value. Use tabs between columns when values contain commas.
Confirm column roles. Dose scale and units are explicit choices below.
Observation details · IDs, units, SD and exclusions
| Column row | Group | Response | Dose | Unit | Replicate ID | SD (σ) | Exclude | Reason |
|---|
Start with your table or a labelled synthetic example.
Analysis of Data 1
Previous calculations
Read-only records. These results use the engine and source data saved at that time. The current calculation and data stay unchanged.
Graph of Data 1
No downloads started yet.Paired rows and omissions
Click the title or an axis label to edit it in Format Graph. Double-click a point to format its data set. All plotted marks and text remain editable in SVG.
Bars show 95% Student t intervals for means, not population spread. Each dot is one observation.
Download history
These records confirm that a download started. Saving or reading the file is not verified. Current means that the content still matches; stale means that it differs. Older records without content evidence stay unknown.
Residuals
Technical analysis record
Exact saved calculation for reproducibility. Read the explanation beside the results for interpretation.
Run an analysis to inspect its record.
Method limits, file support, and privacy
A project contains named data tables, each with its own analysis and graph. Save project includes all tables, their source records, calculations and figure settings. This early version keeps one current analysis and graph per table; Info, Layout and analysis chains are not supported. This app supports independent one-factor groups and one 4PL curve per analysis. It does not validate clinical decisions, technical-replicate independence, or complex study designs. It does not choose a method for you.
Column and XY grids edit the same observation records. Column rows keep their saved table positions. Paired tests use values on the same row; subject IDs are optional metadata. A missing or excluded value omits that pair without shifting later rows. XY Y subcolumns retain each observed replicate at the same X. Metadata remains in Observation details. CSV and TSV: long-form or multi-group Column tables, one header row. Column import retains headers and source row positions, including missing cells and empty rows; an optional subject-ID column is metadata. Extra unselected columns are listed and are not imported. XLSX: choose one worksheet, first row as headers, raw cell values; formulas use their saved result. PZFX: choose one supported numeric Column or XY table. Source row positions, missing responses, Y subcolumn order and Y exclusion flags are retained. Unsupported tables remain separate choices with a reason. Column tables need one Y subcolumn per data set. XY tables need distinct numeric X rows, with at most one missing X. Excluded X values, duplicate or blank data-set names, aggregate mean/error data and other X formats are refused. Original analyses, graphs, styling, layouts, row titles, units, original exclusion reasons and other tables are not transferred. Set concentration units and input scale before dose-response analysis. Read the import report before use. Full .prism import is not supported.
Source data, results, and graphs stay in this browser. The app has no upload endpoint, analytics, or remote calculation service. Save a project file for backup and colleague transfer. Browser storage can be cleared by your browser. No shared live collaboration is claimed.
Offline: the complete runtime is supplied locally and cached after first load. Keep the local server running for a portable localhost entry. Exported Python scripts require Python, NumPy, and SciPy; they include the exact engine source and your data.