<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[singingbell的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/singingbell</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; singingbell]]></copyright><item><title><![CDATA[WIN7/WIN10    MATLAB安装硬件支持/安装PYTHON-OPENCV获取摄像头实时图像]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/115532160</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/115532160</guid><author>singingbell</author><pubDate>Thu, 08 Apr 2021 22:37:18 +0800</pubDate><description><![CDATA[WIN7/WIN10  MATLAB安装硬件支持/安装PYTHON-OPENCV获取摄像头实时图像

1. PYTHON-OPENCV

准备：Anaconda3-5.2.0-Windows-x86_64；opencv_python-4.4.0.40-cp36-cp36m-win_amd64.whl；msgpack-1.0.2-cp36-cp36m-win_amd64.whl

（1）通过Anaconda安装python3.6：

1）双击安装exe，照提示一直安装就好；

2）将此PYTHON3.6设..]]></description><category></category></item><item><title><![CDATA[实现MATLAB循环或一次性读取句柄的小技巧]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/105301472</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/105301472</guid><author>singingbell</author><pubDate>Fri, 03 Apr 2020 21:41:58 +0800</pubDate><description><![CDATA[方法1：
比如GUI界面有四个单选按钮，对应的handles结构变量中就有radiobutton1、radiobutton2、radiobutton3、radiobutton4四个句柄tag

可以先将struct中对应tag的值存成一个变量a，然后再直接引用a或者a(i)得到句柄对应的指针值
a=[handles.radiobutton1,handles.radiobutton2,handles...]]></description><category></category></item><item><title><![CDATA[MATLAB灵活改变画图位置]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/105199695</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/105199695</guid><author>singingbell</author><pubDate>Mon, 30 Mar 2020 16:15:04 +0800</pubDate><description><![CDATA[自定义子图像位置与大小
subplot(‘Position’,positionVector);
%positionVector给定左下角位置与图像宽高[left, bottom, width, height]
y = magic(4);
figure
positionVector1 = [0.1, 0.2, 0.3, 0.3];
subplot('Position',positionVector1...]]></description><category></category></item><item><title><![CDATA[MATLAB解方程]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104855038</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104855038</guid><author>singingbell</author><pubDate>Sat, 14 Mar 2020 16:42:41 +0800</pubDate><description><![CDATA[1. 求解一般的代数方程
例：
在这里插入代码片
syms a b c x
S=a*x^2+b*x+c
solve(S)
求解表达式等于0的根

或再输入一个参数指定未知数b=solve(S,b)
2. 求解线性方程
1） AX=B的解X=A\B或X=inv(A)B
2）XA=B的解X=B/A或X=Binv(A)
3. 求解线性微分方程
r=dsolve(‘eq1,eq2,…’,‘cond1,co...]]></description><category></category></item><item><title><![CDATA[MATLAB特殊绘图函数]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104815592</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104815592</guid><author>singingbell</author><pubDate>Fri, 13 Mar 2020 20:52:32 +0800</pubDate><description><![CDATA[1. 极坐标
polar(theta,rho)极坐标绘图
theta = 0:0.01:2*pi;
rho = sin(2*theta).*cos(2*theta);
figure
polar(theta,rho,'--r')
 
rose(theta)极坐标频数累计柱状图
load sunspot.dat
theta = sunspot(:,2);
figure
rose(theta,12)
 ...]]></description><category></category></item><item><title><![CDATA[回归分析]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104799005</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104799005</guid><author>singingbell</author><pubDate>Wed, 11 Mar 2020 21:27:18 +0800</pubDate><description><![CDATA[一、一元回归分析
主要步骤：
（1）画观测值，观察是否符合线性分布；
（2）若符合线性分布：
a）采用最小二乘法或其他求取参数；
b）采用相关系数检验或F-检验或其他方法验证是否线性假设，结论的显著性；
c）若验证正确解决进一步的预测或控制问题。
（3）若非线性分布，采用二次或多次多项式进行拟合求解参数
1. 一元线性回归
（1）模型

（2）最小二乘法求解参数：

（3）相关系数检验：


（4...]]></description><category></category></item><item><title><![CDATA[三、方差分析+回归分析]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104635729</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104635729</guid><author>singingbell</author><pubDate>Wed, 11 Mar 2020 16:05:56 +0800</pubDate><description><![CDATA[p=anova1(x)
p=anova1(x, group)



]]></description><category></category></item><item><title><![CDATA[二、假设检验]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104598503</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104598503</guid><author>singingbell</author><pubDate>Wed, 11 Mar 2020 16:05:18 +0800</pubDate><description><![CDATA[[h, sig, ci]=ztest(x, m, sigma, alpha, tail)


]]></description><category></category></item><item><title><![CDATA[一、MATLAB参数估计]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104598013</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104598013</guid><author>singingbell</author><pubDate>Sun, 01 Mar 2020 19:35:15 +0800</pubDate><description><![CDATA[1. [N, K]=hist(data, k)
简单统计返回数据data落在每一个区间的频数N和每一个区间的中点X。
2. h=normplot(x)
判断数据是否来自于正态分布，若为正太分布显示出直线性形态，其他显示曲线形态。
%不同分布数据生成
rng(11)  % For reproducibility
x1 = normrnd(0,1,[50,1]);
x2 = trnd(5,[50,1]...]]></description><category></category></item><item><title><![CDATA[word写论文时标题与页面顶端的距离和设定不一样的问题]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/102779407</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/102779407</guid><author>singingbell</author><pubDate>Sun, 01 Mar 2020 18:58:24 +0800</pubDate><description><![CDATA[问题
在使用word写论文的，可能遇到如下图的问题 ，标题的顶端与页眉的距离太近，不是设定的上下段距离。

可以有两种解决办法：


将分页符改为分节符，但是这样会使剩下的页面重新编码，不承接上面的页数；



将



...]]></description><category></category></item><item><title><![CDATA[六、计算元素之间差diff+近似梯度gradient+子空间夹角subspace]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104505329</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104505329</guid><author>singingbell</author><pubDate>Sun, 01 Mar 2020 17:33:02 +0800</pubDate><description><![CDATA[计算元素之间的差diff
近似梯度gradient


]]></description><category></category></item><item><title><![CDATA[五、概率和临界值即随机数的产生]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104505307</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104505307</guid><author>singingbell</author><pubDate>Sat, 29 Feb 2020 14:45:12 +0800</pubDate><description><![CDATA[正太分布随机数rands
均匀分布随机数rand


]]></description><category></category></item><item><title><![CDATA[MATLAB rands.m函数部分注释]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104573790</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104573790</guid><author>singingbell</author><pubDate>Sat, 29 Feb 2020 14:29:43 +0800</pubDate><description><![CDATA[function out1 = rands(in1,in2,in3,in4,in5,in6)
%  BOILERPLATE_START
%  This code is the same for all Weight/Bias Initialization Functions.
persistent INFO;
if isempty(INFO)
  if nnet.test.flags('initW...]]></description><category></category></item><item><title><![CDATA[四、最大max+最小min+平均值mean+元素和sum]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104496154</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104496154</guid><author>singingbell</author><pubDate>Tue, 25 Feb 2020 20:45:10 +0800</pubDate><description><![CDATA[1. 最大值max
（1）M=max(A)
a）输入为一维向量，输出其中的最大值；
b）输入为二维矩阵，以行向量来输入每个列向量的最大值；
c）输入为多维矩阵，沿第一个长度不为1的数组维度做最大化输出，输出该维度的大小为1其他维度大小保持不变（第一个维度为0返回同样尺寸的空矩阵）。
注意若输入的为复数，则判断依据为幅值最大。
（2）M=max(A, [], dim)
对输入矩阵A沿维度dim做最大...]]></description><category></category></item><item><title><![CDATA[三、直方图histogram和条形图bar]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104349744</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104349744</guid><author>singingbell</author><pubDate>Tue, 25 Feb 2020 12:51:23 +0800</pubDate><description><![CDATA[四、直方图和棒图histogram

]]></description><category></category></item><item><title><![CDATA[二、向量的向量积cross+向量点积dot+列元素的积prod]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104335062</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104335062</guid><author>singingbell</author><pubDate>Sun, 16 Feb 2020 21:33:07 +0800</pubDate><description><![CDATA[##二、 向量的向量积cross
（1）C=cross(A,B)
输入A、B必须为长度为3的向量，得到长度为3的与A、B同时垂直的向量C。
过程可用代码解释为：
A=[1 2 3];
B=[4 5 6];
AB=[a;b];
dd(1)=det(ab(:,[2,3]));
dd(2)=det(ab(:,[1,3]));
dd(3)=det(ab(:,[1,2]));
C=[dd(1),dd(2),...]]></description><category></category></item><item><title><![CDATA[一、求相关函数 corrcoef+协方差矩阵cov]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/104248786</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/104248786</guid><author>singingbell</author><pubDate>Sat, 15 Feb 2020 20:32:27 +0800</pubDate><description><![CDATA[求相关函数corrcoef
一般作用：1）结合图表评估数据拟合效果；2）评估两组数据之间的互相关联程度，大于0正相关，小于0负相关，等于0不相关。
但此种计算方法反映的是“线性相关”程度，对非线性相关结果不可靠（可采用斯皮尔曼等级相关系数(Spearman correlation coefficient)来评估两参量之间的“单调相关性”）。
(1) R=corrcoef(A)
例：矩阵A 由3个长...]]></description><category></category></item><item><title><![CDATA[PS笔记]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/103595359</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/103595359</guid><author>singingbell</author><pubDate>Wed, 18 Dec 2019 11:58:14 +0800</pubDate><description><![CDATA[PS笔记

1）一般照片大小和分辨率，300ppi，5x7寸照片画布
2）移动工具+Alt直接复制粘贴不添加图层，ctrlc+ctrlv复制到新图层3
3）Ctrl+alt+shift+t同一图层里复制粘贴且重做变换
4）导航器可以和放大缩小一块儿使用，吸管工具（标尺）可以和信息调板一起使用可以量取角度长度等信息（可以通过旋转摆正图片）
5）图层-普通图层-设置为智能对象，就变为了矢量图形，可以...]]></description><category></category></item><item><title><![CDATA[MATLAB画柱状图对比]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/103589436</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/103589436</guid><author>singingbell</author><pubDate>Tue, 17 Dec 2019 23:07:16 +0800</pubDate><description><![CDATA[对21个类别两个分类方法的分类结果用MATLAB画柱状图对比
#输入两个类别的分类结果a1,a2
#输入类别名字name
y=[a1,a2];#y的size为[19,2]
b=bar(y);#画柱状图
axis([0,24,0,1]);#设定坐标轴的显示范围
grid on;#加网格
hold on;#保持图像
set(gca,'xticklabel',name);
set(gca,'XGrid...]]></description><category></category></item><item><title><![CDATA[正二十面体的各个面位置点]]></title><link>https://blog.youkuaiyun.com/singingbell/article/details/102904420</link><guid>https://blog.youkuaiyun.com/singingbell/article/details/102904420</guid><author>singingbell</author><pubDate>Mon, 04 Nov 2019 20:05:57 +0800</pubDate><description><![CDATA[正二十面体的各个面位置点
MATLAB
clear all;clc;
%------------------------------------------------------
phi = (1+sqrt(5))/2;
vert = [1/phi phi 0;-1/phi phi 0;1/phi -phi 0;-1/phi -phi 0];
vert = [vert;phi 0 1/phi; ...]]></description><category></category></item></channel></rss>