Changelog
graphstatsr 2.6.1
Added Features :
- Bar plots with the same behavior as box plots; more informative for samples without replicates.
- Option to add value labels to each bar in bar plots.
- Download button (PDF and JPEG).
Improvement:
- Simplified archive download for JPEG files.
graphstatsr 2.5.0
Added Features :
- Boxplots (EasyStats) with continuous scale: Boxplots can now be generated using a continuous X scale, which makes it possible to visualize kinetics when needed. (Select Type of variable X: continuous)
- PNG export buttons for EnrC13 (IsoPlot)
- Interactive Plotly output for EnrC13 (IsoPlot)
- Simplified downloaded archives (results are now placed directly at the root of the archive)
Debugs:
- Results from different analysis sessions ending up in the same archive
- Boxplot generation (EasyStats) failing when no data is available in the first metabolite or first condition group
- Boxplot coloring in preview blocking the rest of the process
Unresolved, known issue:
- Bug related to double spaces in metadata: in HTML, consecutive spaces are automatically collapsed into a single one, causing inconsistencies between the original dataset and the information displayed/rendered by the web application.
Conclusion: Avoid using multiple spaces in metadata.
graphstatsr 2.2.0
NEW MODULE
MSPT: submodule of Isoplot, able to analysis “pascal triangle” sample to validate experiment. - reference: Millard, Pierre, et al. 2014. “Isotopic Studies of Metabolic Systems by Mass Spectrometry: Using Pascal’s Triangle To Produce Biological Standards with Fully Controlled Labeling Patterns.” Analytical Chemistry 86 (20): 10288–95. See : - https://doi.org/10.1021/ac502490g; - https://github.com/llegregam/PascalTriangle
graphstatsr 1.10.0
Added Features :
- rbase graphics show faster rendering than ggplot.
- users can apply yaxis minimum / maximum limits and steps size, for ggplots figures and rbase graphics
- boxplot with ggstatsplots, pdf and image outputs
- metadata template based on features table vertical / horizontal display, labels size control for ggplot outputs.
- download all boxplots in png/jpg format in a single- archive
- button to download tests datasets
- widget to output specific features
- conditionnal popping buttons avoiding confusion new button to download all ACP files run / update- ACP with one button.
Fixes:
- metadata filters
- disabling statistics tests when plotting individuals - boxplots.
graphstatsr 1.5.0
Added Features:
- New widget allowing user to sum quantification of compounds belonging the same group
- PCA: combining factors, display ellipses or not
- Boxplots: pngs output for each graphic fixes
graphstatsr 1.4.1
Added Features:
- add new input modules with filters widgets (datamods package) add options for boxplots:
- coloring boxplots,
- custom y axis label
- inform outliers or not
- reordering/stashing boxplot conditions (sortable package)
- modularizing code
graphstatsr 1.3.2
Added Features:
- New widget to export up to four box plots per PDF page.
- Box plots available in interactive Plotly mode.
- Option to export all plots in PNG format, named by metabolite.
- Tooltips to display outlier samples in box plots.
- Labels to identify outlier samples in box plots.
- Improved handling of missing values (NA).
Debugs:
- Fixed NA handling in PCA when some columns contain only NA values.
- Fixed uniqueness of row names in the feature table.
- Proper handling of zero values in the weighting factor (treated as NA).
- Clicking rows in the merged table now allows filtering of outliers.
- Box plots: added an option to display all conditions (even those without values) or hide them.
- Fixed handling of the µ character.