C++ double error - Tilted Forum Project Discussion Community

Rekna · 10-20-2004, 07:44 PM

What does the it mean when a double outputs as these error codes?

-1.#J
-1.#IND

theFez · 10-20-2004, 10:07 PM

-1.#IND is NaN (not a number) sqrt(-1), 0.0f / 0.0f will both give this result.

not sure about the other.

littlebighead · 10-22-2004, 02:07 AM

Never seen -1.#J before. Looks interesting. Are you going to tell us how you got that value?

Rekna · 10-22-2004, 11:42 PM

the -1.#IND was comming from the square root of a negative. I have no clue what caused the -1.#J I'm running some linear algerbra algorithms on some large matrix's to demonstrate floating point errors. So it is probably realted to that somehow.

10-20-2004, 07:44 PM	#1 (permalink)
Rekna Junkie	C++ double error What does the it mean when a double outputs as these error codes? -1.#J -1.#IND

10-20-2004, 10:07 PM	#2 (permalink)
theFez Crazy Location: here and there	-1.#IND is NaN (not a number) sqrt(-1), 0.0f / 0.0f will both give this result. not sure about the other.

10-22-2004, 02:07 AM	#3 (permalink)
littlebighead Upright Location: Berkeley, CA	Never seen -1.#J before. Looks interesting. Are you going to tell us how you got that value?

10-22-2004, 11:42 PM	#4 (permalink)
Rekna Junkie	the -1.#IND was comming from the square root of a negative. I have no clue what caused the -1.#J I'm running some linear algerbra algorithms on some large matrix's to demonstrate floating point errors. So it is probably realted to that somehow.