How can I read an Excel spreadsheet into a Matlab table and not have the storage requirement for the table increase by a factor of 10 over the Excel file size?

I have Excel spreadsheets of size 57MB, 220MB, and 136MB. When I use "readtable" to read these spreadsheets, the table size is 5.1GB, 18.7GB, and 11.4GB, respectively, which are too large to effectively use in Matlab. Obviously, I can just not use the table data structure, but is there a way of reading a spreadsheet into a Matlab table without increasing the storage requirement so drastically?

Here is the code: file = 'S:\IM\CSN Files\FOR DR.SINCOVEC\NCHEMS\dbo_NCHEMS_uno.xlsx' T = readtable(file); save('NCHEMS_UNO_DATA', 'T', '-v7.3')

All these spreadsheets have 53 columns with the largest having about 814,800 rows.

Answer 1

[Peter Perkins]

It's likely that you're reading in text data. Consider using categorical variables and datetimes in the table. readtable has had features added over several releases, so it's hard to give specifi advice without knowing what release you're on. ImportOptions is a recent addition that should help.