How to delete all strings with NaN in Struct

Question

Volker 2023-8-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2003027-how-to-delete-all-strings-with-nan-in-struct

评论： Volker 2023-8-3

Hello,

I've been wasting hours trying to remove the rows in the struct variable ID with value = 'NaN' in osmScenarios.trafficLight.ID.

ID is a string in this case.

For example:

arep = strrep(osmScenarios.trafficLight.ID,'NaN','')

error: Too many input arguments.

I know that probably something is wrong with me using that struct but what the change to happen in that struct. No, I need it to happen in that struct.

Please help. i don´t know what to do anymore.

Best regards

3 个评论
显示 1更早的评论隐藏 1更早的评论

Stephen23 2023-8-1

编辑：Stephen23 2023-8-1

"strrep(osmScenarios.trafficLight.ID,'NaN','') error: Too many input arguments."

Most likely the field TRAFFICLIGHT is non-scalar, thus the dot-indexing TRAFFICLIGHT.ID produces a comma-separated list with multiple arrays. If that is the case, you might need to use a loop+indexing, or perhaps ARRAYFUN... what exactly you need depends on the sizes and arrangement of the arrays stored in that structure, which you have not explained nor uploaded. Not uploading your actual data will significantly slow down your getting a useful solution.

"I've been wasting hours trying to remove the rows in the struct variable ID with value = 'NaN' ..."

It is unclear how STRREP would help you to "remove the rows in the struct variable" because it does not do anything like "remove the rows" of a structure, it simply replaces some characters with some other characters (just as its documentation states). So as well as providing your actual data, you need to explain more clearly what you want to achieve, preferably with both the input data and expected output data.

Volker 2023-8-3

Hi tanks for the support. As I answered in a comment below I was able to solve the problem. Thank you for your time! :)

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Mrutyunjaya Hiremath 2023-8-1

2
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2003027-how-to-delete-all-strings-with-nan-in-struct#answer_1281577

在 MATLAB Online 中打开

Based on your description, it seems like you have a struct variable osmScenarios, and within it, there is a field trafficLight that contains another field ID. The values in the ID field are strings, and you want to remove the rows where the value is 'NaN'. However, the 'NaN' values in strings cannot be directly removed using strrep since it's designed for replacing substrings in strings.
To remove the rows with 'NaN' values in the ID field of the osmScenarios.trafficLight struct, you can do the following:

% Assuming osmScenarios is the struct variable
for i = numel(osmScenarios.trafficLight):-1:1
    if strcmp(osmScenarios.trafficLight(i).ID, 'NaN')
        osmScenarios.trafficLight(i) = [];
    end
end

Here's what the code does:

We loop through the elements of osmScenarios.trafficLight in reverse order (from end to start) using numel to get the number of elements. This is done to ensure that the removal of elements does not affect the index of subsequent elements.
We check if the ID field of the current element is equal to 'NaN' using strcmp.
If the condition is true, we remove the element from the osmScenarios.trafficLight struct using indexing (osmScenarios.trafficLight(i) = []).

After running this code, the rows with 'NaN' in the ID field will be removed from the osmScenarios.trafficLight struct.
Note: If the 'NaN' values are actual NaN numeric values instead of strings, you can use the isnan function to identify and remove those rows. But based on your description, it seems like you are dealing with string values of 'NaN'.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Volker 2023-8-3

Thank you for your support! I actually had this soulution but as a newby in programming I didn´t think about using a reversed loop. I my case a loop from from 1:... doesn´t work because after deleting an element the next element would change index. With reversing the loop I solved this problem. Your answer is correct though since I didn´t mention this in my question :)

请先登录，再进行评论。