Changelog
0.1.3
heatmap:interpolation="nearest"for deterministic cell rendering across matplotlib versionsREADME: full rewrite — install, 3 examples (raincloud, ROC, UpSet), badges, Colab badgeexamples/demo.ipynb: outputs regenerated forlazy-visual==0.1.3- bump to 0.1.3
0.1.2
heatmap: RdBu_r default, robustvmax, grid off over data, larger annotation fontsupset:A (n)labels inside bars, Y-axis on the right, grid off, fixed indentationraincloud: geometric layout (half-violin + box + jittered dots) without overlapbox: adaptive p-value bracket spacing, wider caps, smaller fliersviolin: black quartiles/median (no red default), no duplicate bracket textscatter: 95% CI band for linear fit, legend outside axesdistribution: transparent hist, colored KDE lines, external legendroc: DeLong viascipy.rankdata, thinned plot for fast rendering, legend outsidecore/style.py: disabled minor ticks,grid.which = "major"only_common.py: grid only whenshow_grid=True, notight_layoutfor multi-panel (upset) or colorbar (heatmap)- bump to 0.1.2
0.1.1
LvResult.show(): no-op in Jupyter inline backends (avoid double figure)render/mpl.py: removed forcedAggbackend so Jupyter/Colab auto-displaysscatter/roc/distribution: legend outside axes withouttight_layoutwarnings- bump to 0.1.1
0.1.0
First public release.
- 8 plot functions:
scatterplot,distribution,box,violin,raincloud,heatmap,upset,roc - built-in stats: Mann-Whitney U, Welch t-test, Kruskal-Wallis, Holm correction, Cohen's d, DeLong & bootstrap AUC CI
- 4 themes:
default,nature,apa,dark - export: PNG / SVG / PDF (atomic writes)
lv.load()for CSV, TSV, Parquet, Excel, JSON, HTTP(S) URLLvResultwithfig,ax,spec,stats,tables,files- 22 error / warning codes (
LV-E101..E310,LV-W501..W504)