how do I move mixed (numeric, categorical, logical} data into a table for subsequent analysis

Question

Richard Palmer 2017-2-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/326019-how-do-i-move-mixed-numeric-categorical-logical-data-into-a-table-for-subsequent-analysis

评论： Richard Palmer 2017-2-23

I am using the adult incomes dataset. It contains numeric and nominal data. It also contains a nominal/logical variable (did income exceed 50k?). fitglm works with a table object. How to I get the appropriate column flags into a table for analysis?

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

Nachiket Katakkar 2017-2-23

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/326019-how-do-i-move-mixed-numeric-categorical-logical-data-into-a-table-for-subsequent-analysis#answer_256041

在 MATLAB Online 中打开

An example of moving mixed data types into a single table followed by calling "fitglm" is shown below. This uses the "carsmall" data-set within MATLAB but should be easy to extend to the data you have mentioned:

% Load the carsmall dataset
load carsmall
% Convert values to nominal and logical
Cylinders = nominal(Cylinders);
Weight = Weight > 3000;
% Make a table
tbl = table(Cylinders,Weight,Origin,MPG);
% Fit GLM model
mdl = fitglm(tbl);

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Richard Palmer 2017-2-23

Thanks! Do you know how to get from the carsmall to a dataset that can be processed by the neural net packages. I setup a tbl like yours "by hand" and got fitglm to work. However, I don't know how to get the tbl into a form amenable for neural net analysis. Since I have a lot of nominal data in my dataset, I need a way to mechanize this.

Again, thanks!

请先登录，再进行评论。