科研插件
数学与统计学
Mathematics & Statistics
面向数学与统计研究:贝叶斯统计、因果推断、时间序列、生存分析、拓扑数据分析、数值线性代数、优化与运筹、统计功效。
官方收录发布者 Codar安全审核通过v0.1.115 个 Skill
README
数学与统计学(Mathematics & Statistics)
面向数学与统计研究:贝叶斯统计、因果推断、时间序列、生存分析、拓扑数据分析、数值线性代数、优化与运筹、统计功效。
来源与许可
本 Plugin 中的 Skill 精选自以下开源仓库,保留原许可;每个 SKILL.md 的 source 字段指向具体出处。
- KalarisLabs/research-agent-skills — MIT,1 个 Skill
- K-Dense-AI/scientific-agent-skills — MIT,6 个 Skill
- xjtulyc/awesome-rosetta-skills — MIT,8 个 Skill
示例提示词
请用 PyMC 为这份分层数据建立贝叶斯多层模型,并做后验预测检验。
包含的科研 Skills15 个
- 贝叶斯统计建模Builds Bayesian models with PyMC and ArviZ: priors, NUTS sampling, convergence diagnostics, posterior checks, LOO-CV, hierarchical models. Use for Bayesian inference workflows.v0.1.0
- 因果推断工具包Estimates causal effects with DoWhy/CausalML: DAG construction, backdoor identification, ATE estimation, propensity matching, refutation tests. Use for intervention-effect questions.v0.1.0
- 网络图分析与社区发现Builds and analyzes graphs with networkx/igraph: centrality, shortest paths, community detection, random graph models. Use for network structure and connectivity analysis.v0.1.0
- 数值线性代数计算Performs SVD, PCA, eigendecomposition, Cholesky, iterative solvers, sparse matrices, and condition-number analysis with numpy/scipy. Use for numerical-stability questions.v0.1.0
- 运筹学优化求解Formulates and solves LP/QP/MILP with cvxpy and OR-Tools: sensitivity analysis, dual variables, solver selection. Use for operations-research and resource-allocation problems.v0.1.0
- PyMC贝叶斯概率编程Builds and fits Bayesian models in PyMC: hierarchical models, NUTS/MCMC, variational inference, LOO/WAIC comparison. Use for probabilistic-programming inference workflows.v0.1.0
- 多目标进化优化Solves multi-objective optimization with pymoo: NSGA-II/III, MOEA/D, Pareto fronts, constraint handling, ZDT/DTLZ benchmarks. Use for conflicting-objective engineering design.v0.1.0
- 离散事件仿真建模Builds discrete-event simulations in SimPy with processes, queues, and shared resources over time. Use for manufacturing, service, logistics, or network system models.v0.1.0
- 统计检验方法选择与报告Guides hypothesis-test selection, assumption checks, power analysis, and APA-formatted reporting. Use for academic statistical reporting and test-choice guidance.v0.1.0
- 统计功效与样本量计算Calculates sample size and power for study planning via closed-form and Monte Carlo methods, covering t-tests, ANOVA, regression, mixed models. Use to justify a sample size for a grant or IRB.v0.1.0
- statsmodels统计建模Fits statistical models (OLS, GLM, mixed models, ARIMA) in statsmodels with detailed diagnostics and coefficient inference. Use for econometrics and rigorous model-based inference.v0.1.0
- 生存分析方法Analyzes time-to-event data with lifelines: Kaplan-Meier, log-rank test, Cox PH diagnostics, AFT models, competing risks. Use for survival or time-to-event outcomes.v0.1.0
- SymPy符号数学计算Performs exact symbolic math in Python: algebra, calculus, equation solving, symbolic linear algebra, LaTeX/code generation. Use when exact, not floating-point, results are needed.v0.1.0
- 时间序列预测分析Analyzes time series end to end: stationarity tests, ACF/PACF, ARIMA/SARIMA, STL decomposition, VAR, GARCH. Use for forecasting and temporal-dependency modeling.v0.1.0
- 拓扑数据分析Extracts topological features via persistent homology, Betti numbers, persistence diagrams, and Mapper graphs. Use to find multi-scale shape structure classical stats miss.v0.1.0
常见问题
数学与统计学能做什么?
面向数学与统计研究:贝叶斯统计、因果推断、时间序列、生存分析、拓扑数据分析、数值线性代数、优化与运筹、统计功效。
如何安装数学与统计学?
在科搭 Codar 左侧点击「插件」,在「插件」页签搜索「数学与统计学」并安装;也可以把本页的安装提示词发给 Claude Code 或 Codex。
数学与统计学需要哪些权限?
该科研插件没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
数学与统计学经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。