Issue with importing excel files

I posted a question similar to this, but I did not realize what was really going on until recently.

My excel file has multiple rows of the following form.

f 27//1 29//2 5//3 2//4

For this line, it is separated into 5 cells, which I will indicate with | signs.

f | 27//1 | 29//2 | 5//3 | 2//4

When I want to call the cell with 27//1 in it, i put..

txt(1,2)

and it gives me the following.

'27//1'

I need to extract the number to the left of the // sign. The problem is, I believe the single quotation marks from the import are messing things up. When I actually TYPE in 27//1 in the regexp function it works.

str2double(regexp('27//1','\d+(?=//)','match'))

However when I call the cell in the regexp it does NOT. I get NaN

str2double(regexp(txt(1,2),'\d+(?=//)','match'))

How can I circumvent this? I tried to search for a number between a ' and a // sign, but it didn't work.

str2double(regexp(txt(1,2),'(?<=")\d+(?=//)','match'))

Answer 1

[Geoff Hayes]

Matthew - the problem isn't with the single quotes in your cell input, but with the str2double. Since you are passing a cell array (with only one string) as your input to regexp, the return value from that will be a cell array. For example,

 >> var = {'27//1'};
 >> regexp(var,'\d+(?=//)','match')

 ans = 
    {1x1 cell}

If we look closer at ans we see that its value is

 >> ans{1} 

 ans = 
    '27'

And since this answer is not a string but a cell, then the str2double returns NaN.

There are couple of ways to get around this:

1. Pass the string from the cell array into the regexp using curly braces: str2double(regexp(txt{1,2},'\d+(?=//)','match'))
2. Convert the cell element to a string: str2double(regexp(char(txt(1,2)),'\d+(?=//)','match'))

Try out the first method and see what happens!

[Matthew Brandsema]

Yes! The first one worked! Thank you for your reply and more importantly your description as to what is going on!