Return to DrawPlotia

DrawPlotia 0.1.0

Finding projects after a website address change

Saved figures belong to the browser profile and website address where you created them. If you used the previous website, open it in the same browser profile, open your saved figures and use Save project to download each portable .obf.json file. Then open those files in DrawPlotia at the new address. Keep your downloaded backups and the old site's browser data until you have checked the imported figures. Changing the app's name does not change the portable project format.

Create and edit scientific figures locally, without an account or uploads.

Create and edit

Choose New figure or a starter template. Add text, shapes and arrows from the toolbar. Double-click text to edit. Drag objects and resize with their handles. Shift-click selects several objects; the Style inspector edits position, size, color and rotation.

Expand Scientific drawing in the library for cells, membranes, DNA, panels, scale bars and editable charts. Browse 583 drawings across 33 categories, filter by topic or license, and use Show more to load additional results. CC BY artwork requires attribution: download credits from License and keep them with PNG/PDF publications. Check each illustration's scientific suitability. Scale-bar values and example chart data need author review.

Save and recover

Edits autosave in this browser. Wait for Saved locally. Download a portable project backup with Save project file; its artwork and provenance travel with it. Browser storage belongs to this address and port, and can be cleared or exhausted. If a save fails, download the project before closing.

The Figures tab in the editor's left sidebar lists every figure saved in this browser, newest first. Select one to reopen it; the open figure's edits are saved first. Delete removes a figure's saved copy and its Recent projects entry after confirmation; downloaded project files are not affected. Saved figures stay only in this browser and are never uploaded. Clearing the site data removes them. If local storage is full, delete saved figures you no longer need or download a project backup.

Export

SVG preserves vectors and live text. PNG exports document pixels at the chosen 1–4× scale. PDF opens the browser print dialog: choose Save as PDF, 100% scale, no margins and no headers/footers. Chrome/Edge honor the figure page size at 96 pixels per inch.

Data, plots and statistics

Suggested types appear at the top of the right sidebar; click a type to apply it. Dot plot with cross bar shows observations and a horizontal aggregate marker. Dot plot with bar/column overlays observations on an aggregate column. Bar / column is an ordinary column chart. Grouped bar requires row data with separate X-group, Series and Value columns; series appear side by side within each X group. Column-header means alone are ordinary columns.

For a spreadsheet with one measurement group per column, choose Data arrangement → Column names on X, or Plot column means. Selected column headers label X; the initial chart shows each column’s mean with SEM error bars. Uncheck numeric IDs and unwanted columns. Missing cells are excluded per column, never replaced by zero. The preview lists each mean and its valid-value count. Change to line, area or another style while keeping the selected columns. Group and value columns uses row-wise groups; Replicate columns within groups pools selected measurements by a group column.

Insert plot has three steps: Choose data, Review data, and Design plot. Back and the earlier step buttons keep your draft while you navigate. Reset clears the draft and returns to the beginning, preserving data and plots already saved in the figure. New data and its plot are saved together only when you Apply; one Undo removes both.

Choose Insert plot in the toolbar, or open the Data tab beside Figures. Import a CSV, TSV or text file, paste spreadsheet cells, or choose Type values. Check the delimiter, header, decimal/thousands separators, missing tokens, date format and column types before adding the dataset. Ambiguous numbers require confirmation. The preview shows the first 150 rows with excluded-cell counts; all rows enter the selected calculations.

Choose a plot type and map its columns explicitly. There are 15 types: bar/column, grouped bar, stacked bar, line, area, range, scatter, heatmap, box, violin, dot plot with cross bar, dot plot with bar/column, grouped dot plot, Cleveland and Manhattan. Duplicate line positions and heatmap cells need an explicit aggregate. Missing values are excluded and reported. A box always shows its Tukey outliers; a violin with one value or zero variance shows observations instead of an invented density. Manhattan p = 0 is reported and capped at −log10(p) = 300.

No statistical test is selected automatically. One-way ANOVA compares all plotted independent group means and optionally adds Tukey HSD (Tukey–Kramer for unequal sample sizes), with adjusted p-values and simultaneous confidence intervals. It assumes equal variances and approximately normal residuals. The overall F test and comparison brackets can be shown independently. Each X × series combination is one group; this is not a two-way or repeated-measures ANOVA.

Choose specific pairs with Group A, Group B and Add comparison, all pairs, or a control. Grouped data also offer comparisons within each X or series. Choices persist when switching tests. Welch, Student, paired t, Mann–Whitney and Wilcoxon compare the chosen pairs; Kruskal–Wallis optionally adds Dunn comparisons. Correction uses the planned family, including failed comparisons. Tukey always uses the complete all-pairs family, even if only selected comparisons are displayed, without an additional correction. Hiding brackets never changes p-values. Pairwise tests are two-sided.

Paired tests require unique subject IDs within each group, checked before missing measurements are excluded. Complete matches and excluded observations are reported per comparison. Column-header groups can explicitly match the same source row; missing cells never shift later matches. Replicate columns cannot use row-order pairing. Group summaries include all valid observations; paired tests use the matched subset. The results table and spreadsheet-safe CSV/Copy table report the same calculations as the plot.

A plot is one canvas object. Resize it, edit its labels in Style, or choose Edit plot to change mappings and analysis. Source data and plot settings travel with the portable project; results are recomputed on open. Replacing a dataset preserves mapped columns by name. A missing or incompatible column displays an error frame until repaired. Delete dataset refuses data used by a plot.

Everything runs locally, including parsing and analysis workers. Limits are 5 MB per text source, 20 MB per figure, 20 sources, 50 datasets, 100,000 rows, 500 columns and 1,000,000 padded cells. Figures also have aggregate vector-export limits; reduce points, labels or plots when refused. Original-source downloads preserve the imported text, except its leading Unicode BOM.

Données, graphiques et statistiques

Les types suggérés sont en haut du panneau droit. Choisissez explicitement un type. Points avec barre transversale montre les observations et un repère horizontal ; Points avec barre / colonne les superpose à une colonne. Les barres groupées nécessitent des colonnes distinctes de groupe X, série et valeur dans des données par ligne. Les séries apparaissent côte à côte dans chaque groupe X.

Pour un groupe de mesures par colonne, choisissez Disposition des données → Noms de colonnes en X, ou Tracer les moyennes des colonnes. Les en-têtes définissent X ; le graphique initial montre les moyennes et leurs erreurs standard. Décochez les identifiants numériques et les colonnes inutiles. Chaque moyenne utilise ses valeurs valides, sans remplacer les cellules manquantes par zéro. L’aperçu indique les moyennes et leurs effectifs. Le style du graphique reste modifiable.

Le parcours comporte trois étapes : Choisir les données, Vérifier les données et Créer le graphique. Retour et les étapes précédentes conservent le brouillon. Réinitialiser efface le brouillon et revient au début, sans supprimer les données et graphiques déjà enregistrés. Les nouvelles données et le graphique sont enregistrés ensemble à l’application ; une seule annulation retire les deux.

Choisissez Insérer un graphique dans la barre d’outils ou ouvrez l’onglet Données. Importez un fichier CSV, TSV ou texte, collez des cellules de tableur ou saisissez des valeurs. Vérifiez le séparateur, l’en-tête, les formats numériques et de dates, les valeurs manquantes et les types de colonnes. Confirmez explicitement les nombres ambigus. L’aperçu affiche les 150 premières lignes ; tous les calculs utilisent l’ensemble des lignes.

Choisissez explicitement un test. L’ANOVA à un facteur compare tous les groupes indépendants tracés, avec HSD de Tukey (Tukey–Kramer pour des effectifs différents) ou sans comparaisons post-hoc. Elle suppose des variances égales et des résidus approximativement normaux. Le test F global, les valeurs p ajustées, les intervalles simultanés et les annotations sont disponibles. Chaque combinaison X × série est un groupe : ce n’est pas une ANOVA à deux facteurs ni à mesures répétées.

Choisissez des paires précises avec Groupe A, Groupe B et Ajouter une comparaison, toutes les paires ou un témoin. Les choix persistent entre les tests. Welch, Student, t apparié, Mann–Whitney et Wilcoxon testent les paires choisies ; Kruskal–Wallis peut ajouter Dunn. La correction inclut les comparaisons non calculables. Tukey conserve toujours la famille complète, même si seules certaines paires sont affichées. Masquer les annotations ne modifie pas les valeurs p. Les tests par paires sont bilatéraux.

Les tests appariés exigent des identifiants uniques dans chaque groupe, vérifiés avant l’exclusion des mesures manquantes. Les effectifs appariés et exclus sont indiqués par comparaison. Les groupes en colonnes peuvent explicitement utiliser la même ligne source ; les cellules manquantes ne décalent pas les paires. Les colonnes de répétitions ne permettent pas cet appariement. Les résumés utilisent toutes les mesures valides, les tests appariés seulement les sujets correspondants. La table, le CSV et le graphique présentent les mêmes calculs.

Le graphique reste un objet unique, redimensionnable et modifiable. Les données et les réglages accompagnent le fichier de projet. Les résultats sont recalculés à l’ouverture. Une colonne remplacée incompatible affiche un cadre d’erreur jusqu’à correction. Tout fonctionne localement, sans envoi de données.

Rights and safe imports

The License panel shows rights records and exports credit/publication reports. Private SVG imports stay unreviewed. Supply known creator, source and license information; the app does not infer permission. Unsupported SVG effects, scripts and external resources are rejected before insertion.

Shortcuts

Local distribution

Keep the local server window open. The built folder runs without internet with Node.js 24.11 or newer. Copy the complete distribution and portable projects to another computer for offline use. This release has no service-worker installation or automatic update system.

Code: Apache-2.0, adapted from OpenBioFigure. Artwork: pinned CC0 and CC BY 4.0 records with per-item credits. Fonts and dependencies retain their separate notices in third-party licenses.