Minor changes:
Minor changes:
plotCorrelation function.Minor changes:
plotCounts, plotCountsPerBiotype: Fixed unwanted type mismatch during
internal leftJoin operations for SingleCellExperiment method.Minor changes:
plotCountsCorrelation: Add back xName and yName arguments, which
are useful for customizing the plot labels.Major changes:
plotPCA to plotPca. Now using AcidGenerics variant instead of
BiocGenerics here.plotQC to plotQc.plotTSNE to plotTsne.plotUMAP to plotUmap.Major changes:
acid_, for consistency.acid_coord_flip as acid_discrete_coord_flip. This no longer
attempts to parse the gg object, and returns ggproto, so it can be
chained in + operations.snake_case conventions:
autoContinuousColorScale to acid_scale_color_continuous;
autoContinuousFillScale to acid_scale_fill_continuous;
autoDiscreteColorScale to acid_scale_color_discrete;
autoDiscreteFillScale to acid_scale_fill_discrete.Minor changes:
Major changes:
.data) instead of legacy
syntactic sugar approach with !! and sym. Refer to rlang documentation
for details on non-standard evaluation and tidy evaluation.Minor changes:
linewidth instead
of size for line geoms.plotCountsPerCell: Hardened point mode for SCE objects that don't contain
precalculated nCount metadata in colData. Also improved error message
when user attempts to plot point with a geom other than ecdf.plotCorrelation: Now using recommended after_stat instead of deprecated
stat function for correlation calculations.DFrame instead of DataFrame virtual class.requireNamespaces import from AcidBase to goalie.Minor changes:
facet_wrap by using vars consistently.
This requires calling sym with !! or syms with !!! (from rlang).Minor changes:
Minor changes:
Major changes:
Minor changes:
as_tibble coercion).Major changes:
plotUpset: Now using ComplexUpset engine interanlly.Minor changes:
matchLabels: Simplified argument matching using new argName approach
instead of legacy choices approach. Also improved hardening against
mismatched user input, when attempting to specify elements in labels list
that are not allowed/supported internally in the function.Minor changes:
matchLabels: Now allowing input of logical(1), so we can better support
dynamic automatic plot labels.Minor changes:
acid_geom_label_repel: Now setting max.overlaps = Inf internally, to
ensure that all points are always labeled. Refer to ggrepel examples
documentation for details on max.overlaps argument, and the override
option ggrepel.max.overlaps, which works globally in an R session.Minor changes:
geoms.Rd file.guides() handling to use character string instead of
logical (e.g. FALSE), which was changed in ggplot 3.3.5+.Major changes:
plotCorrelation: Updated data.frame method to support labelCol, which
is useful for defining the column to use for plot labels.plotCorrelation: The trans now defaults to "log10", instead of "identity".
This is recommended for IC50 plots, which are the intended use of this
function the majority of the time.Minor changes:
Minor changes:
Minor changes:
percent from scales, which is used in pointillism.Minor changes:
!!, !!!, sym, and syms.Matrix::colSums internally. Just inherit from the
regular S4 generic.Minor changes:
draw_plot,
ggdraw, log_breaks, plot_grid, and pretty_breaks.New functions:
plotCorrelation: Quick X-Y correlation scatterplot that calculates the R2
value, similar to plots in GraphPad Prism.Minor changes:
plotWaterfall: Reworked internal code to better handle splitting out of
interesting groups into separate plots.SummarizedExperiment method support to plotWaterfall.sampleID internally as sample
identifier column, helping improve migrate to sampleId as preferred method.Minor changes:
plotCounts: Added an assert to check for user attempting to use bar geom
with style = "wide". Also added sort argument, which allows the user
to easily plot the genes alphabetically instead of by vector order.Minor changes:
plotCounts: Added geom argument to support additional visualization types
other than dots. Can now plot boxplot, violin, or bar plot.Minor changes:
plotHeatmap: Improved internal factor and NA handling of pheatmap annotation
metadata, which can otherwise cause some legends to not render correctly.
Note that we're ensuring factors get releveled correctly here.Major changes:
plotUpset: reworked default formals to now include nIntersects and
orderBySize arguments. Improved internal argument handling and assert
checks, ensuring user is passing in an integer matrix containing only 0,1
values. Improved the data.frame method to drop non-0,1 columns automatically,
which supports the UpSetR movies working example.plotPCA: Reworked subtitle and title formals to instead use the
standardized labels approach instead. Now automatically plots the number
of genes used, which defaults to the top 500 most variable.Minor changes:
color and fill arguments to inherit better from ggplot2
defaults: scale_colour_continuous and scale_colour_discrete. The defaults
still inherit from acid.color.continuous and acid.color.discrete first but
then fall back to the ggplot2 default global options
ggplot2.continuous.colour and ggplot2.discrete.colour.Minor changes:
Minor changes:
plotUpset: Adjusted ratio of dot matrix to main barplot, to handle complex
multiple comparisons better.Minor changes:
plotUpset: Added matrix and list method support. Simplified internal code,
reducing dependencies from UpSetR. Maybe consider switching to ggupset in a
future update.Minor changes:
plotWaterfall: Now supports log transformations, defaulting to log10.
Labels are disabled by default, and now only supported for non-log plots.
For log transformed data, a line is now shown at the y intercept.New functions:
plotWaterfall: Easily plot concentration curve (e.g. IC50) waterfall
plots.Minor changes:
matchLabels now returns an empty list for a NULL argument. This approach
is used in the upcoming pointillism update, in the plotStackedBarPlot
function.Minor changes:
plotCounts, plotHeatmap, plotQuantileHeatmap: Added
convertGenesToSymbols support to override automatic mapping of genes to
symbols, which is enabled by default.plotHeatmap, plotQuantileHeatmap: Improved showRownames handling for
automatic visualization of rows, which are now enabled by default for heatmaps
with <= 50 features.Minor changes:
plotHeatmap unit tests to harden against SummarizedExperiment and
SingleCellExperiment objects defined in AcidTest package that contain
rows and/or columns with all zeros. Refer to nonzeroRowsAndCols function
in basejump for details.Minor changes:
acid_coord_flip: Fix for breaking change introduced by ggplot2 v3.3.0,
which renamed object[["layers"]][[1L]][["geom"]][["required_aes"]] from
"x" to "x|y". The package now requires this new version of ggplot2.Minor changes:
aggregateCellsToSamples instead of
pseudobulk, which has migrated to pointillism and will be reworked in an
upcoming release.SingleCellExperiment methods, where
aggregateCellsToSamples is called internally. This was propagating to other
packages where SingleCellExperiment methods are not defined, making the
documentation confusing.Minor changes:
Minor changes:
plotHeatmap: Increased the default break range from -2:2 up to -3:3.
This helps improve the dynamic range a bit for some RNA-seq plots.Minor changes:
interestingGroups input against NA values causing
values to drop in plots. This was discovered with plotPCA calling
geom_point internally, which caused removal of rows. If we coerce NA value
to "NA" first, then removal no longer occurs. This is accomplished
internally using str_remove_na (see also stringi_remove_na). The change
applies to these functions: plotCellCounts, plotCounts,
plotCountsPerBiotype, plotCountsPerBroadClass, plotCountsPerFeature,
plotFeaturesDetected, plotPCA, plotTotalCounts, plotZerosVsDepth.Minor changes:
plotCounts: The gene argument is now optional, supporting subset
SummarizedExperiment objects. Currently a maximum of 20 genes (rows) are
allowed to pass through in the method.plotHeatmap, plotCorrelationHeatmap,
and plotQuantileHeatmap.Minor changes:
plotCounts: Reworked medianLine argument and renamed to simply line.
Now supports mean and geometric mean in addition to median now.
Line is now disabled by default.Minor changes:
matchLabels in documentation.Major changes:
title argument with labels argument for functions that render
plots using ggplot2.matchLabels function, that handles internal ggplot2 label
matching, and auto-populates with defaults in formal argument.Minor changes:
SingleCellExperiment S4 methods to
SummarizedExperiment, using standard ... passthrough instead of relying
upon matchArgsToDoCall.plotCounts: Moved DESeqDataSet method to DESeqAnalysis package.Major changes:
acid_theme_dracula: New dark mode theme supporting Dracula color palette.lightPalette, darkPalette,
draculaPalette, iOSLightPalette, iOSDarkPalette, macOSLightPalette,
macOSDarkPalette.synesthesia: Simplified color palette to use purple, blue, green, orange.
User can tweak the palette using a new palette argument.Minor changes:
plotCountsPerFeature: Renamed formals to match conventions in new basejump
melt methods.Minor changes:
plotPCA: Improved automatic ntop argument handling when passing in a
value greater than the number or features (rows) defined in the object.SingleCellExperiment methods that need to average
or aggregate expression at sample level.Minor changes:
Major changes:
Minor changes:
Minor changes:
Minor changes:
pheatmap calls.makeLabel and
makeTitle functions from syntactic update.New functions:
plotCountsCorrelation and plotCountsCorrelationHeatmap: New functions
optimized for plotting correlation between gene expression matrices.
Particularly useful for comparing pseudoaligned vs. aligned counts, which is
implemented in the new bcbioRNASeq package update.Minor changes:
Minor changes:
plotCountsPerBiotype: Internal tidyeval bug fix for call to dplyr::top_n.
Previously, the n argument supported direct variable input, but now it
must be unquoted, due to change in underlying rlang engine following the
0.4 update.Minor changes:
upset: Compatibility update for UpSetR v1.4 release.Minor changes:
plotCountsPerBiotype and
plotCountsPerBroadClass. Previously, the breaks argument was defined in
internal scale_y_continuous ggplot2 call, but this doesn't perform well for
many RNA-seq datasets at log2 scale. Instead, it's better to log2 scale the
data first and then simplify indicate log2 on the y axis title.plotCountsPerFeature: Improved custom color handling for density geom.
Removed violin geom support but added jitter geom support. The violin
plot method doesn't scale well for multiple samples, and the boxplot is
visually easier to interpret for this metric, in general.Major changes:
Major changes:
upset function from UpSetR, with modified formals, so we can
easily call this from other packages, including DESeqAnalysis.New functions:
gradient palette functions: purpleOrange, blueYellow.Major changes:
plotHeatmap now supports automatic color calculation when breaks argument
is defined. Also improved support for legendBreaks.Minor changes:
.pheatmapColorPalette internal check for color, which requires n
argument, rather than assuming positional variable.plotCountsPerFeature: Improve the count subtitle.Reworked package, consolidating previous minimalism and firestarter packages.
New functions:
synesthesia color palette functions.acid_coord_flip: Intelligently puts samples at the top of the Y axis when
flipped. Particularly useful for barplots.acid_pretty_breaks: Attempt to improve scales::pretty_breaks, using an
approach more similar to the base plot engine.acid_theme_light, acid_theme_dark.Major changes:
ggproto functions) now consistently
use acid_ as a prefix.Minor changes:
plotCountsPerBiotype and plotCountsPerBroadClass now support title.Deprecations:
plotCountsPerGene has been renamed to plotCountsPerFeature.plotGenesDetected has been renamed to plotFeaturesDetected.theme_paperwhite and theme_midnight in favor of
acid_theme_light and acid_theme_dark.New functions:
base_breaks: An attempt as automatic pretty breaks based on the default R
plotting engine. Derived from a Stack Overflow post (see documentation).Minor changes:
plotCountsPerBiotype, plotCountsPerBroadClass switched to recommending
log10 counts by default.Updated packages depencies and switched to Docker for Travis CI.
Initial release, migrating ggplot2 functions from basejump package.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.