《Origin8-数据处理》由会员分享，可在线阅读，更多相关《Origin8-数据处理（67页珍藏版）》请在金锄头文库上搜索。
1、Department of Chemistry and Life Science, University of Chuzhou. *数据处理 (1)：数学运算Department of Chemistry and Life Science, University of Chuzhou. *数学运算菜单n插值与外推n简单数学运算n归一化n微分n积分n平均Department of Chemistry and Life Science, University of Chuzhou. *一维插值步骤n工作表读入数据文件（Mathematicsinterpolation.dat)n既可以从工作表选中数
2、据进行一维插值，也可以将选 中数据作图后进行一维插值n选择菜单命令 Analysis: Mathematics: Interpolate/Extropolate Y from Xn设置插值对话框有关选项Department of Chemistry and Life Science, University of Chuzhou. *一维插值Department of Chemistry and Life Science, University of Chuzhou. *一维插值Department of Chemistry and Life Science, University of Chu
3、zhou. *Trace Interpolate(对环形或周期曲线)Department of Chemistry and Life Science, University of Chuzhou. *三维插值步骤n导入数据，如： SamplesMathematics3D Interpolation.datn激活工作表，选择菜单 Analysis : Mathematics : 3D Interpolationn指定X、Y、Z、F，和每维多少数据点（如5，即差值 出125个点）n按“OK”Department of Chemistry and Life Science, University o
4、f Chuzhou. *三维插值-设置输入和控制参数Department of Chemistry and Life Science, University of Chuzhou. *绘制3D散点图Plot : 3D XYZ : 3D ScatterDepartment of Chemistry and Life Science, University of Chuzhou. *绘制3D散点图-设置点大小Department of Chemistry and Life Science, University of Chuzhou. *导数运算-Mathematics: Differentiat
5、eDepartment of Chemistry and Life Science, University of Chuzhou. *数据处理 (2):回归与拟合*Department of Chemistry and Life Science, University of Chuzhou. *什么是回归分析 n所谓回归(regression)分析，简单的说，就是一种处理 变量与变量之间相互关系的数理统计方法。用这种数 学方法可以从大量观测的散点数据中寻找到能反映事 物内部的一些统计规律，并可以按数学模型形式表达 出来，故称它为回归方程（回归模型）。 n例如，自由落体运动中，物体下落的距离S与
6、所需时 间t之间，有如下关系：S = gt2/2n变量S的值随t而定(其他项是常数)，这就是说，如果t 有确定值，那么S的值就完全确定了。这种关系就是 所谓的函数关系或确定性关系。n回归(Regression)也可以称为拟合(Fitting)，回归是要 找到一个有效的关系，拟合则要找到一个最佳的匹配 方程，两者基本是同一个意思。Department of Chemistry and Life Science, University of Chuzhou. *回归分析分类 n根据方程涉及变量的个数u一元回归 u多元回归(multiple regression) n根据自变量和因变量函数关系是直线
7、还是曲线u线性回归(linear regression) u非线性回归 (nonlinear regression) u(非线性回归处理的情况要比线性回归复杂得多，需要进行 更大量的尝试。因此除了依赖计算进行反复运算逼近，用户 自己对参数的取值范围和估算也很重要。)Department of Chemistry and Life Science, University of Chuzhou. *主要拟合方式nFit Linear: 一元线性拟合nFit Polynomial: 一元多项式拟合nMultiple Linear Regression: 多元线性回归nNonLinear Curve
8、Fit: 非线性曲线拟合nNonlinear Surface Fit: 非线性表面拟合nSimulate Curve:查看曲线（可修改参数）nSimulate Surface:查看曲面（可修改参数）nFit Exponential:指数拟合nFit Sigmoidal:S型曲线拟合Department of Chemistry and Life Science, University of Chuzhou. *线性拟合分析 nOriginLabOrigin8SamplesCurve Fitting Linear Fit.dat 。nAnalysis Fitting Fit Linear Dep
9、artment of Chemistry and Life Science, University of Chuzhou. *拟合结果的分析报表 Department of Chemistry and Life Science, University of Chuzhou. *排除部分数据点拟合nSamplesCurve FittingOutlier.datDepartment of Chemistry and Life Science, University of Chuzhou. *查看各数据误差n第6个数据标准差太大Department of Chemistry and Life Sci
10、ence, University of Chuzhou. *排除数据点n在激活的图上加入排除点(按住左键不放调出菜单)Department of Chemistry and Life Science, University of Chuzhou. *多项式回归 nOriginLabOrigin8SamplesCurve Fitting Polynomial Fit.dat Department of Chemistry and Life Science, University of Chuzhou. *多元线性回归n导入文件Multiple Linear Regression.datnAnal
11、ysis-Fitting-Mulitple Linear Regressionn设置D列为Dependent Data，A,B,C列为Independent DataDepartment of Chemistry and Life Science, University of Chuzhou. *非线性拟合 n设置函数Department of Chemistry and Life Science, University of Chuzhou. *设置拟合数据Department of Chemistry and Life Science, University of Chuzhou. *设置
12、拟合和报表形式Department of Chemistry and Life Science, University of Chuzhou. *拟合结果Department of Chemistry and Life Science, University of Chuzhou. *转化为线性拟合n公式n在工作表增加两列数据（D、E）n设D为Xn分别设D和E列为A和B列倒数Department of Chemistry and Life Science, University of Chuzhou. *线性拟合n对D和E数据作图n可以进行Fit Linear或Fit PolynomialDe
13、partment of Chemistry and Life Science, University of Chuzhou. *修改参数选择n下部数据拟合不好，可以重新选择数据Department of Chemistry and Life Science, University of Chuzhou. *选择下部数据区Department of Chemistry and Life Science, University of Chuzhou. *用户自定义函数拟合n菜单Tools:Fitting Function OrganizerDepartment of Chemistry and L
14、ife Science, University of Chuzhou. *步骤n按NewCategory-重命名User-Defined按New Function改 名为MyExp设置parameter names和FunctionDepartment of Chemistry and Life Science, University of Chuzhou. *编译函数n按Function旁边按钮 调出Code Builder窗口按 Compile，然后按Return to Dialog按钮返回。返回窗口 按Save再按OK退出Department of Chemistry and Life
15、Science, University of Chuzhou. *使用nCurve FittingExponential Decay.datn选中B列。打开非线性拟合对话框，选择自定义函数 。然后在参数里将y0,a,b依次设置80，100，-5。Department of Chemistry and Life Science, University of Chuzhou. *非线性曲面拟合n导入Matrix Conversion and GriddingXYZ Random Gaussian.datn将C列设置为z轴n菜单plot-3D xyz-3D Scatter做3D图n菜单analysi
16、s-fitting-Nolinear Surface Fitn对话框中设置拟合函数为Gauss2Dn按Fit按钮进行拟合操作Department of Chemistry and Life Science, University of Chuzhou. *非线性曲面拟合Department of Chemistry and Life Science, University of Chuzhou. *数据处理 (3):统计分析Department of Chemistry and Life Science, University of Chuzhou. *相关系数n导入文件correlations.dat。选择B和C列，执行菜单命 令Correlation Coefficient，在对话框plot中勾中scatter plotsDepartment of Chemistry and Life Science, University of Chuzhou. *描述性统计-statistics on columnsn文
