Why do I get NaN values?

Question

Stelios Fanourakis 2019-4-22

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/457878-why-do-i-get-nan-values

评论： Walter Roberson 2019-4-24

I have used exactly the same code for Sheet 5 and different column names and I get numerical values but when I use it at Sheet 6 and slightly adjacent columns but the same number of rows I get NaN.

Any idea why this happens?

x1 = readtable('ValidationTest.xls', 'Sheet',6, 'Range','N2:N385 ');
y1 = readtable('ValidationTest.xls', 'Sheet',6, 'Range','A2:A385 ');
 
x2 = readtable('ValidationTest.xls', 'Sheet',6, 'Range','O2:O868');
y2 = readtable('ValidationTest.xls', 'Sheet',6, 'Range','B2:B868');
 
 
P1 = polyfit(x1{:,:}, y1{:,:}, 6 ) 
P2 = polyfit(x2{:,:}, y2{:,:}, 6)
 
allX = unique([x1{:,:}; x2{:,:}]);
 
Pd = polyval(P1, allX) - polyval(P2, allX)
 
P11 = polyval(P1, allX)
P22 = polyval(P2, allX)
 
hold on
plot(allX, Pd, 'b')
plot(allX, P11, 'g')
plot(allX, P22, 'r')

13 个评论
显示 11更早的评论隐藏 11更早的评论

Stelios Fanourakis 2019-4-23

在 MATLAB Online 中打开

In that case

%% Setup the Import Options
opts = spreadsheetImportOptions("NumVariables", 15);
 
% Specify sheet and range
opts.Sheet = "Sheet6";
opts.DataRange = "A2:O873";
 
% Specify column names and types
opts.VariableNames = ["m2readypng", "A6algpng", "Var3", "Var4", "Var5", "Var6", "Var7", "Var8", "Var9", "Var10", "Var11", "Var12", "Var13", "XX", "XXX"];
opts.SelectedVariableNames = ["m2readypng", "A6algpng", "XX", "XXX"];
opts.VariableTypes = ["double", "double", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "double", "double"];
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "WhitespaceRule", "preserve");
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "EmptyFieldRule", "auto");
 
% Import the data
ValidationTestS6 = readtable("/Users/steliosphanourakes/Desktop/Matlab/DesMoines/karadokei - Copy/ValidationTest.xls", opts, "UseExcel", false);
 
%% Convert to output type
e = table2array(ValidationTestS6);
 
%% Clear temporary variables
clear  opts
 
P1 = polyfit(e(:,4), e(:,2), 6 ) 
P2 = polyfit(e(:,1), e(:,3), 6)
 
allX = unique([e(:,3); e(:,4)]);
 
Pd = polyval(P1, allX) - polyval(P2, allX)
 
P11 = polyval(P1, allX)
P22 = polyval(P2, allX)
 
hold on
plot(allX, Pd, 'b')
plot(allX, P11, 'g')
plot(allX, P22, 'r')

I used the automated Matlab method to import from excel.

Until the P1 = polyfit command the data are imported fine as numbers. Can you spot now the difference and why my import method didn't work?? (my method is described at the begining of this thread).

Now using the polyfit command I start getting NaN from that point and onwards. Why is this happening? Does it matter that the two columns I am trying to polyfit are not the same size? Why size should matter?

Stelios Fanourakis 2019-4-24

在 MATLAB Online 中打开

I finally made it run with numbers but I don't think the results are correct since I am not getting plots in the end

{[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.0230]}
    {[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.0660]}
    {[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.1090]}
    {[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.1520]}
    {[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.1950]}
    {[   NaN]}    {[5.6760]}    {[    NaN]}    {[37.2380]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
    {[   NaN]}    {[     0]}    {[    NaN]}    {[    NaN]}
Warning: Polynomial is not unique; degree >= number of data points. 
> In polyfit (line 74)
  In see (line 33) 
P1 =
   1.0e-03 *
    0.3916         0         0         0         0         0         0
Warning: Polynomial is not unique; degree >= number of data points. 
> In polyfit (line 74)
  In see (line 34) 
P2 =
   1.0e-03 *
    0.3916         0         0         0         0         0         0
Pd =
     0
P11 =
    4.8020
P22 =
    4.8020

Is it correct if I write the commands with the brackets like this?

I want to say the 4th whole column of the cell matrix with the 2nd whole column of the cell matrix as X,Y and apply the 6th degree polynomial. The same underneath for 3rd and st column as X,Y. Is it correct written?

P1 = polyfit(e{4}, e{2}, 6 ) 
P2 = polyfit(e{3}, e{1}, 6)
 
allX = unique([e{3}; e{4}]);

Stelios Fanourakis 2019-4-24

在 MATLAB Online 中打开

In the case of that code

clc;
clear all;
 
%% Setup the Import Options
opts = spreadsheetImportOptions("NumVariables", 15);
 
% Specify sheet and range
opts.Sheet = "Sheet6";
opts.DataRange = "A2:O873";
 
% Specify column names and types
opts.VariableNames = ["m2readypng", "A6algpng", "Var3", "Var4", "Var5", "Var6", "Var7", "Var8", "Var9", "Var10", "Var11", "Var12", "Var13", "XX", "XXX"];
opts.SelectedVariableNames = ["m2readypng", "A6algpng", "XX", "XXX"];
opts.VariableTypes = ["double", "double", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "double", "double"];
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "WhitespaceRule", "preserve");
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13], "EmptyFieldRule", "auto");
 
% Import the data
ValidationTestS6 = readtable("/Users/steliosphanourakes/Desktop/Matlab/DesMoines/karadokei - Copy/ValidationTest.xls", opts, "UseExcel", false);
 
%% Convert to output type
e = table2array(ValidationTestS6);
 
 
x1 = e(:,4)
y1 = e(:,2)
x2 = e(:,3)
y2 = e(:,1)
 
 
 
 
%% Clear temporary variables
clear  opts
 
P1 = polyfit(x1, y1, 6) 
P2 = polyfit(x2, y2, 6)
 
allX = unique([x1; x2]);
 
Pd = polyval(P1, allX) - polyval(P2, allX)
 
P11 = polyval(P1, allX)
P22 = polyval(P2, allX)
 
hold on
plot(allX, Pd, 'b')
plot(allX, P11, 'g')
plot(allX, P22, 'r')

I get numbers until allX. From Pd and onwards I get NaN.

Why is that?

Stelios Fanourakis 2019-4-24

在 MATLAB Online 中打开

Finally, after using that code I get results and plots. The point is that not for all excel sheets and datasets I get the exactly same graph, as in the excel. The graph I get in Matlab differs the graph I get from excel for some sheets.

Can you take a look at the script please, and tell me if anything is wrong?

%% Import data from spreadsheet
% Script for importing data from the following spreadsheet:
%
%    Workbook: /Users/steliosphanourakes/Desktop/Matlab/DesMoines/karadokei - Copy/ValidationTest.xls
%    Worksheet: Sheet12
%
% Auto-generated by MATLAB on 24-Apr-2019 17:55:58
 
clc;
clear all;
 
%% Setup the Import Options
opts = spreadsheetImportOptions("NumVariables", 16);
 
% Specify sheet and range
opts.Sheet = "Sheet12";
opts.DataRange = "A2:P868";
 
% Specify column names and types
opts.VariableNames = ["m8readypng", "A36algpng", "Var3", "Var4", "Var5", "Var6", "Var7", "Var8", "Var9", "Var10", "Var11", "Var12", "Var13", "XX", "Var15", "XXX"];
opts.SelectedVariableNames = ["m8readypng", "A36algpng", "XX", "XXX"];
opts.VariableTypes = ["double", "double", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "string", "double", "string", "double"];
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15], "WhitespaceRule", "preserve");
opts = setvaropts(opts, [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15], "EmptyFieldRule", "auto");
 
% Setup rules for import
opts.ImportErrorRule = "omitrow";
opts.MissingRule = "omitrow";
opts = setvaropts(opts, [1, 2, 3, 4], "TreatAsMissing", '');
 
% Import the data
ValidationTestS12 = readtable("/Users/steliosphanourakes/Desktop/Matlab/DesMoines/karadokei - Copy/ValidationTest.xls", opts, "UseExcel", false);
 
%% Convert to output type
ValidationTestS12 = table2array(ValidationTestS12);
 
%% Clear temporary variables
clear  opts
 
e = ValidationTestS12;
 
x1 = e(:,4)
y1 = e(:,2)
x2 = e(:,3)
y2 = e(:,1)
 
%after the readtable
mask = isnan(x1) | isnan(y1);
x1(mask,:) = [];
y1(mask,:) = [];
mask = isnan(x2) | isnan(y2);
x2(mask,:) = [];
y2(mask,:) = [];
 
 
P1 = polyfit(x1, y1, 6)
P2 = polyfit(x2, y2, 6)
 
allX = unique([x1; x2]);
 
Pd = polyval(P1, allX) - polyval(P2, allX)
 
P11 = polyval(P1, allX)
P22 = polyval(P2, allX)
 
hold on
plot(allX, Pd, 'b')
plot(allX, P11, 'g')
plot(allX, P22, 'r')

Stelios Fanourakis 2019-4-24

编辑：Stelios Fanourakis 2019-4-24

@John

The graph is correct. It should be like this more or less. By luck, I sent you one dataset with a constant value. Not all sheets are the same. It’s just happened for the particular one. Most of the datasets are curves.

Hopefully, I need to ask as many questions as I need in order to get the right answer. Hope it won’t bother.

Walter Roberson 2019-4-24

That code is not set up to read the VTest.xlsx file you provided.

The earlier code was set up to read it (once the file name was changed), but it did not take into account that two of the columns were shorter than the others.

请先登录，再进行评论。

请先登录，再回答此问题。