Android可视化统计模块

本文汇总了截至2012年3月的Android平台上的图表绘制库和工具,包括开源和商业选项。涵盖ChartDroid、AndroidPlot、AChartEngine等知名库的特点及使用场景。

ince this question was asked a few very nice open source native libraries have been created. These include:

  1. ChartDroid
  2. AndroidPlot
  3. AChartEngine

And a commercial chart:

  1. ShinobiControls

ORIGINAL

It looks like the creator of this question solved this problem and posted some interesting stuff about it to the Internet. I've linked to his solution along with several others that exist now.

  1. rapidandroid (question creator solution)
  2. GraphView
  3. Java Charts for Android

Just to summarize: his solution involved using a JavaScript library (flot) built on top of jQuery. This library was then included in an HTML page which was loaded into a custom WebView. The custom WebView then read data from a custom Java class made accessible to JavaScript code via the WebView.addJavascriptInterface method. Details can be found at the above link.

The other solutions may present a more straightforward approach.



Android Charts Graphs Diagrams plots plotting charting

This is my personal compilation of Android charting and plotting libraries and tools as of March 2012, I might have used some of the previous comments on this page / stackoverflow without referencing, for which I am very thankful.

  1. Opensource libraries
  2. Commercial libraries
  3. Tutorial

Due to my low reputation, I cannot post links here, too bad.

Afreechart

http://code.google.com/p/afreechart/

AFreeChart is based on JFreeChart 1.0.13. (C) Copyright 2010,2011 by ICOMSYSTECH Co.,Ltd.

Screenshots:

http://afreechart.googlecode.com/svn/doc/screenshot/index.html

Javadocs:

http://afreechart.googlecode.com/svn/doc/javadocs/index.html

Download:

http://code.google.com/p/afreechart/downloads/list

ShinobiControls

http://www.shinobicontrols.com/android/shinobicharts/product-tour/

ShinobiControls for Android is a commercial chart ($195) with GPU accelerated graphics and multi-touch support.

Droidcharts

http://code.google.com/p/droidcharts/

Attempt to port JFreeCharts to Android platform. Currently the following chart types are supported: pie chart line chart xy line chart bar chart category bar chart

Download:

svn checkout http://droidcharts.googlecode.com/svn/trunk/ droidcharts-read-only

ChartDroid

http://code.google.com/p/chartdroid/

native chart engine for android

ChartDroid is an Intent-based "library application" for static chart and graph generation on Android. It can graph/plot/display numerical data in many representations. As a developer, you need only to have your users install the library, then implement a Content Provider according to this simple specification.

License: Apache License 2.0

Screenshots:

http://code.google.com/p/chartdroid/wiki/Screenshots

AndroidPlot

http://androidplot.com/wiki/Home

License: It's free. See our BSD License androidplot.com/wiki/License

Why use AndroidPlot instead of another 3rdParty library? Dynamic model support Plots auto adjust for screen size and orientation. Actively developed and maintained Responsive support forum

AChartEngine

http://www.achartengine.org/index.html
http://code.google.com/p/achartengine/

It's the only one I've found that is all Java (no webview stuff), is open source, and the author isn't asking for money even if you end up using it in a pay app.

License: Apache License 2.0

Download:

http://code.google.com/p/achartengine/downloads/list

Screenshots:

http://www.achartengine.org/content/demo.html

Tutorial:

http://sites.google.com/site/androidhowto/how-to-1/create-graphs

GraphView

http://android.arnodenhond.com/components/graphview http://github.com/jjoe64/GraphView

Download link:

http://android.arnodenhond.com/components

you are required to build an about page that references your usage of Graphview, and if you charge for your app you have to give a donation to the Graphview author.

Snowdon

http://code.google.com/p/snowdon/

A simple, fast, graphing library for Android featuring scatter plots, line graphs, area plots, histograms, bar charts and heat maps. There is a quick start project in the source, it seems to me that the labels of axes cannot be trivially changed?

rapidandroid

http://github.com/dimagi/rapidandroid

last activity in 2010 (03/2012)

aiCharts for Android

http://www.artfulbits.com/products/android/aicharts.aspx

$150 License

Dynamic Datasource Support

Android Chart

http://www.keepedge.com/products/android_charting/

License: Commercial

$299 License

Java Charts for Android

http://www.java4less.com/charts/chart.php?info=android

costs over $140 USD for a redistribution license so that you can use it on an app that you place on the Android Market.

Rchart 2 for android

http://www.java4less.com/charts/chart.php?info=android

$138.60 License

AndroCharts

http://www.niftymobileapps.com/android-chart-library

$99 License

Another Java - Android only Library with good looking but limited charts. Obvious choice if you want to quickly embed good looking charts with very little coding effort.

Tutorial on Generating graphs in Android

http://sudarmuthu.com/blog/generating-graphs-in-android

goal was to create graphs in android using any JavaScript or HTML that are stored locally. The app and the graphs should work even if access to Internet is not available.

【数据驱动】【航空航天结构的高效损伤检测技术】一种数据驱动的结构健康监测(SHM)方法,用于进行原位评估结构健康状态,即损伤位置和程度,在其中利用了选定位置的引导式兰姆波响应(Matlab代码实现)内容概要:本文介绍了一种基于数据驱动的结构健康监测(SHM)方法,利用选定位置的引导式兰姆波响应对航空航天等领域的结构进行原位损伤检测,实现对损伤位置程度的精确评估,相关方法通过Matlab代码实现,具有较强的工程应用价值。文中还提到了该技术在无人机、水下机器人、太阳能系统、四轴飞行器等多个工程领域的交叉应用,展示了其在复杂系统状态监测故障诊断中的广泛适用性。此外,文档列举了大量基于Matlab/Simulink的科研仿真资源,涵盖信号处理、路径规划、机器学习、电力系统优化等多个方向,构成一个综合性科研技术支持体系。; 适合人群:具备一定Matlab编程基础,从事航空航天、结构工程、智能制造、自动化等相关领域研究的研究生、科研人员及工程技术人员。; 使用场景及目标:①用于航空航天结构、无人机机体等关键部件的实时健康监测早期损伤识别;②结合兰姆波信号分析数据驱动模型,提升复杂工程系统的故障诊断精度可靠性;③为科研项目提供Matlab仿真支持,加速算法验证系统开发。; 阅读建议:建议读者结合文档提供的Matlab代码实例,深入理解兰姆波信号处理损伤识别算法的实现流程,同时可参考文中列出的多种技术案例进行横向拓展学习,强化综合科研能力。
【无人机论文复现】空地多无人平台协同路径规划技术研究(Matlab代码实现)内容概要:本文围绕“空地多无人平台协同路径规划技术”的研究展开,重点在于通过Matlab代码实现对该技术的论文复现。文中详细探讨了多无人平台(如无人机地面车辆)在复杂环境下的协同路径规划问题,涉及三维空间路径规划、动态避障、任务分配协同控制等关键技术,结合智能优化算法(如改进粒子群算法、遗传算法、RRT等)进行路径求解优化,旨在提升多平台系统的协作效率任务执行能力。同时,文档列举了大量相关研究主题,涵盖无人机控制、路径规划、多智能体协同、信号处理、电力系统等多个交叉领域,展示了该方向的技术广度深度。; 适合人群:具备一定Matlab编程基础和路径规划背景的研究生、科研人员及从事无人机、智能交通、自动化等相关领域的工程技术人员。; 使用场景及目标:①用于学术论文复现,帮助理解空地协同路径规划的核心算法实现细节;②支撑科研项目开发,提供多平台协同控制路径优化的技术参考;③作为教学案例,辅助讲授智能优化算法在无人系统中的实际应用。; 阅读建议:建议结合提供的Matlab代码进行实践操作,重点关注算法实现流程参数设置,同时可参照文中列出的其他相关研究方向拓展技术视野,建议按目录顺序系统学习,并充分利用网盘资源进行仿真验证。
### 微调 Qwen3-32B 模型的 LLaMA-Factory 配置流程 在 LLaMA-Factory 中微调 Qwen3-32B 模型,首先需要确保模型和训练框架的兼容性。LLaMA-Factory 在 5.1 版本中已实现对 Qwen3 的深度优化,包括训练和推理逻辑的支持,因此可直接使用该版本进行配置 [^1]。 #### 数据集配置 在微调过程中,数据集的格式字段映射是关键。需要编写符合 LLaMA-Factory 要求的 `dataset_info.json` 文件,并将其放置在容器挂载的 `/datasets` 根目录下。例如,对于采用 ShareGPT 格式的多轮对话数据集,其配置如下: ```json { "Light-R1-SFT-stage2": { "file_name": "qihoo360/Light-R1-SFTData/stage2-3k.json", "file_sha1": "481cd356262d36b9d16ac49f7fc8ff3d4c9f349c", "formatting": "sharegpt", "columns": { "messages": "conversations" }, "tags": { "role_tag": "from", "content_tag": "value", "user_tag": "user", "assistant_tag": "assistant" }, "ranking": false, "field": "auto" } } ``` 该配置确保模型可以正确读取对话结构,并在训练过程中进行有效的上下文理解响应生成 [^4]。 #### 训练配置 训练过程中,需使用 LoRA(Low-Rank Adaptation)模块进行参数高效微调。LLaMA-Factory 支持通过 YAML 配置文件指定训练参数。例如,复制并重命名 `examples/inference/llama3_lora_sft.yaml` 文件,修改其内容如下: ```yaml model_name_or_path: Qwen/Qwen3-32B dataset: Light-R1-SFT-stage2 output_dir: /output/qwen3-32B-lora per_device_train_batch_size: 1 gradient_accumulation_steps: 8 learning_rate: 1e-4 num_train_epochs: 3 logging_steps: 10 save_steps: 500 save_total_limit: 2 lora_rank: 64 lora_alpha: 128 lora_dropout: 0.1 ``` 该配置文件定义了模型路径、训练数据集、输出路径、批量大小、学习率、LoRA 参数等关键参数 [^3]。 #### 启动训练 完成上述配置后,执行训练脚本启动微调流程: ```bash cd LLaMA-Factory python src/train.py --config /path/to/your_config.yaml ``` 若使用多 GPU 环境(如两个 3090 显卡),可在命令中加入分布式训练参数: ```bash torchrun --nproc_per_node=2 src/train.py --config /path/to/your_config.yaml ``` 训练过程中,若显存不足(如训练 32B 模型),可能需要调整 `per_device_train_batch_size` 或启用梯度检查点(gradient checkpointing)等优化策略 [^3]。 --- ###
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值