Originally posted by: HendrikWijnant

Dear all, 

I recently started using CPLEX optimization studio for solving an MILP. While trying to read some float values from excel, I stumbled upon a problem. The problem is that CPLEX apparently changes some of these float values by a factor of 100 and leaves other unchanged. I tried to find why CPLEX does this on the internet or in the CPLEX and OPL guides but I could not find the answer anywhere. 
So what happens is the following:
Using Sheetread, I read some float values into a 2dimensional array of floats. for example values like:

0,0033333
0,0000625
4,1666700
8,3333300
0,0001563

CPLEX apparently reads these values like this:
 0,0033333
0,0000625
416,66700
833,33300
0,0001563

The values > 1 are multiplied by 100 and I have absolutely no idea why. 

This is my code:

In the mod file:

range rRows = 1..nRows;
range rColumns = 1..9;
float data[rRows][rColumns] = ...;

In the data file:

SheetConnection sheet2("Data_50i.xlsx");

I added the datafile in the attachments (it concerns column H, named as v[f])

 Is there anyone that has an idea on how this is possible?

Hi,

if I change your .mod into

int nRows=...;
range rRows = 1..nRows;
range rColumns = 1..9;
float data[rRows][rColumns] = ...;

execute
{
writeln(data[65][8]);

}

I get

4.167

so why do you think you get 416 ?

Which cplex version do you use ?

regards

Originally posted by: HendrikWijnant

Thanks a lot Alex for helping me.
I just noticed that I uploaded the wrong excel file. Could you try it with this file. 

Code .mod:

Code .dat

 SheetConnection sheet("Data_50i.xlsx");

 nRows from SheetRead(sheet,"nRows");
 data from SheetRead(sheet,"Data");

Hi,

in test.csv I get

4.16667

regards

Originally posted by: HendrikWijnant

I found my problem. I opened the csv file using excel and excel apparently imported it incorrect.
Thanks for your help Alex, I greatly appreciate it.