Bokeh 0.9.2 发布，此版本更新内容如下： * Several nan-related fixes including the slow rendering of plots * Removed some unused dependencies * Fixes in our automated release process * Fixed the patchs vanishing on selection * More control over ticks and gridlines * MPL compatibility updated * Several examples updated * bugfixes: - #735 Inconsistent conversion of np.nan to json data in bokehjs - #1005 Nan in data source column causes problem with glyph interface - #1039 Bokeh server can display an empty document without any visible errors - #1075 [tests] Test failure if websocket client is not installed - #1139 Hover tool swaps sides at an off-center position - #1176 Session.store_document() fails in table_server example - #1264 Stock app fails to refresh after drop down is updated - #1381 Fill_color argument doesn't handle rgb(a) tuples properly - #2513 Fix release script failures - #2514 Bug: all patches vanish on selection - #2524 Setting a fixed font size value as a string is deprecated warnings - #2529 [docs] Internal server error on quickstart - #2582 [bokehjs] [regression] Slow plot rendering for 0.9.1 - #2586 [bokehjs] Decimated glyph needs visuals set also - #2593 [docs] Fix typo in charts user guide - #2600 [bokehjs] Hotfix/fixed ticker * features: - #194 [widgets] Widget/controls integration - #2379 Get a warning about an invalid column name - #2496 Improve mouseover information in texas example - #2548 [bokehjs] Need more control over ticks and gridlines * tasks: - #2441 Warning when instantiating plot with no arguments - #2540 Add console warning when bokeh-plot fails - #2541 [docs] Breaking out inline plot examples in user guide into files - #2543 Mpl update - #2546 Clean up examples, use standard bokeh.io output - #2547 [docs] Fix typo in components deprecation message - #2554 [branches] Fix bug on ie (avoid using indices) - #2577 [docs] Editing up to concepts - #2598 Serializing data with numpy optional - #2605 Hotfixes 0.9.2 下载：https://github.com/bokeh/bokeh/archive/0.9.2.zip。 Bokeh (Bokeh.js) 是一个 Python 交互式可视化库，支持现代化 Web 浏览器，提供非常完美的展示功能。Bokeh 的目标是使用 D3.js 样式提供优雅，简洁新颖的图形化风格，同时提供大型数据集的高性能交互功能。Boken 可以快速的创建交互式的绘图，仪表盘和数据应用。 Bokeh 0.9.2 发布，Python 交互式可视化库下载地址
