How to normalize negative values in column vector

Question

Raady 2016-8-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/299913-how-to-normalize-negative-values-in-column-vector

评论： Image Analyst 2016-8-17

Matlab 2015a

Hello, I have a column vector of 180 elements, with min(vector) = -1.6010 and max(vector) = 0.3894. when I apply normc(vector), and find min(vector) = -0.1318 and max(vetor) = 0.0321 for which I except the values to be in between 0 and 1. How to resolve this ??

When I am having positive values I do for ranging between 0 to 1

normalized_vector = (vector -min(vector)).max(vector);

but its not working with the negative values.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Image Analyst 2016-8-17

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/299913-how-to-normalize-negative-values-in-column-vector#answer_232059

在 MATLAB Online 中打开

If you have the Image Processing Toolbox, you can simply use mat2gray:

normalizedVector = mat2gray(vector);

Now normalizedVector will go from 0 to 1. Otherwise, if you are so unfortunate as to not have one of the best toolboxes out there, just scale as you normally would:

normalizedVector = (vector - min(vector)) / (max(vector)-min(vector));

2 个评论
显示无隐藏无

Raady 2016-8-17

编辑：Raady 2016-8-17

Speaking about a vector having all positive values alone. I need to apply this feature to one of the classifier. Which way normalizing data is better using matlab inbuilt function 'normc(vector)' or normalizing to range 0 to 1 by using 'mat2gray'? when I apply these functions output of both are completely different. Please suggest.

Please comment if I need to open as another thread !

Image Analyst 2016-8-17

I don't know. normc() is not a built in function for any of the toolboxes that I have so I don't know what it does.

Both of the statements I give you will give the same results. Both map the min to 0 and the max to 1 and it's linearly scaled in between. I think mat2gray() is the simplest, but you can use whichever one you want.

请先登录，再进行评论。