Find Header in .CSV File and Read Column

I have a large number of .CSV files in one folder. These files have different columns of data. I need to locate the desired columns by matching the header(string) and read only these columns. Example:

Date       Time         Temp_F   Temp_C   Lights_ON   A/C_ON  
5/23/15    12:23:07     70       21       0           1
5/23/15    12:23:08     71       22       0           1
5/23/15    12:23:09     71       22       1           1

Date       Time         Temp_F   Temp_C   Temp_K      Lights_ON   A/C_ON  
5/24/15    13:30:27     63       17       290         1           0
5/24/15    13:30:28     64       18       291         0           0
5/24/15    13:30:29     65       18       291         1           0

I'd like to read Time, Temp_C and Lights_ON from these two files.

Before, the data was the same in all the files. I knew which columns to read and which columns to skip and was using the following code:

c = textscan('%*s%s%*d%d%d%*d', 'Delimiter', ',', 'Headerlines', 1);

Now I need to sort through the headers and match the strings and only read those columns. How would I go about doing this? Can I use textscan to accomplish this? Or maybe csvread or another command? I have attached a sample .CSV file. Thanks in advance.

Answer 1

[Star Strider]

This is one way to approach it:

[n,s,r] = xlsread('Jacee Johnson Book1.csv');
des_cols = {'Time', 'Temp_C', 'Lights_ON'};
colhdrs = s(1,:);
[~,ia] = intersect(colhdrs, des_cols);
colnrs = flipud(ia);
Data = n(:, colnrs-1);

It first reads the file in, then compares the column headers to your desired strings to find the desired columns in ‘colnrs’. Since ‘n’ here is a double array, it then selects the desired columns by numbers, and returns them in ‘Data’. The ‘n’ variable does not include the date (a string), so I subtracted 1 from those columns so they would match appropriately. The time (in Data(:,1) here) is in fractions of a day.

[Walter Roberson]

I think you wanted [~,ia] as your output, not [~ia]

[Star Strider]

Corrected now. Did a copy-paste from working code. Don’t know how the comma got dropped.

[Jacee Johnson]

Thank you very much! This method seems like it will work!

[Star Strider]

My pleasure!

Answer 2

[Walter Roberson]

Consider using the relatively new readtable() as that allows you to address content by column name.

[Jacee Johnson]

I was reading about it but unfortunately I only have access to MATLAB 2011. Thanks for the help.