CODEWITHSUNDEEP
matlabcsv
JB11
JB11

Reputation: 53

What causes matlab to read a number from a CSV file as -INF?

I am trying to read in CSV files to Matlab. For some reason Matlab is converting some of the numbers to -INF.

I have tried using readmatrix and textscan functions. Textscan forces me to use str2double to convert it to a double and it produces the same -INF values as importing the data in with the readmatrix function. Is there a better function to read in a CSV file into MATLAB?

Example: -1.79769313486235E308 is being converted to -INF.

The CSV files are formatted as follows: (with no spaces between the rows)

Header 1, Header 2,           Header 3

190,      -1.47457265853882,  0.0837953791871611

191,      -1.1680064201355,   0.126324314114211

192,      -1.64672422409058,  0.254686301760985

193,      -1.84958028793335,  0.172680441425988

Upvotes: 0

Views: 70

Answers (1)

Wolfie
Wolfie

Reputation: 30047

Within machine precision, -1.79769313486235E308 is -Inf!

From the docs:

MATLAB constructs the double data type according to IEEE® Standard 754 for double precision. The range for a negative number of type double is between -1.79769 x 10^308 and -2.22507 x 10^-308.

So your value here is the most negative value possible for a double in MATLAB. It's likely that Excel uses the same IEEE standard, and is therefore also bottoming out at -Inf just without formatting it accordingly.

Upvotes: 2

Related Questions