北大心理学系.数据使用资料.ppt

数据分析-使用matlab2015-4-14;Matlab statistic toolbox (white book);内容;;Excel数据导入;调用格式： num = xlsread(filename) num = xlsread(filename, -1) num = xlsread(filename, sheet) num = xlsread(filename, range) num = xlsread(filename, sheet, range) num = xlsread(filename, sheet, range, basic) num = xlsread(filename, …, functionhandle) [num, txt]= xlsread(filename, …) [num, txt, raw] = xlsread(filename, …) [num, txt, raw, X] = xlsread(filename, …, functionhandle) xlsread filename sheet range basic;读入文本文件的数据;;;*;*;*;*;;*;*;*;*;*;*;*;*;数据修剪;初步整理;Movies = [ ]; %New Movies variable. Initializing it as empty. M1 = xlsread(Matrix1.xls) %Importing data for Matrix I M2 = xlsread(Matrix2.xls) %Importing data for Matrix II temp = [M1 M2]; %Create a joint temporary Matrix, consisting of two long vectors k = 1; %Initializing index k as 1 for i = 1:length(temp) %Could have said 1603, this is flexible. Start i loop if isnan(temp(i,1)) == 0 isnan(temp(i,2)) == 0 %If both are numbers (=valid) Movies(k,:) = temp(i,:); %Fill with valid entries only k = k + 1; %Update k index only in this case end %End if clause end %End for loop;索引;;;集中趋势;离散度;双变量与多变量描述统计;分类统计-Grpstats;分类统计-Grpstats （参见simonbaseline.m程序）;参数估计与假设检验;数理统计-简单回顾;;;顺序统计量;三个常用分布;区间估计;区间估计;区间估计;区间估计;两正态总体方差比的置信区间;常见分布的参数估计;% 定义样本观测值向量 x = [15.14 14.81 15.11 15.26 15.08 15.17 15.12 14.95 15.05 14.87]; % 调用normfit函数求正态总体参数的最大似然估计和置信区间 % 返回总体均值的最大似然估计muhat和90%置信区间muci， % 还返回总体标准差的最大似然估计sigmahat和90%置信区间sigmaci [muhat,sigmahat,muci,sigmaci] = normfit(x,0.1);正态总体参数的检验;一、总体标准差已知时的单个正态总体均值的U检验; x = [97 102 105 112 99 103 102 94 100 95 105 98 102 100 103]; % 调用ztest函数作总体均值的双侧检验， % 返回变量h，检验的p值，均值的置信区间muci，检验统计量的观测值zval [h,p,muci,zval] = ztest(x,100,2,0.05) % 调用ztest函数作总体均值的单侧检验 [h,p,muci,zval] = ztest(x,100,2,0.05,right);二、总体标准差未知时的单个正态总体均值的t检验;% 定义样本观测值向量 x = [49.4 50.5 50.7 51.7 49.8 47.9 49.2 51.4 48.9]; % 调用ttest函数作总体均值的双侧检验， % 返回变量h，检验的p值，均值的置信区间muci，结构体变量stats [h,p,muci,stats] = ttest(x,50,0.05);三、总体标准差未知