double类型数据初始化

看ffmpeg代码，但到有把double类型的数据初始化为NAN，百度一下，NAN： Not A  Number。

NAN定义在#include <math.h>。

下面直接帖代码介绍一下NAN的应用

	double d = NAN;

	printf("d = %f\n", d);
	if (isnan(d))
	{
		printf("d is NAN\n");
	}
	else
	{
		printf("d is not NAN\n");
	}


	d += 0.05;
	printf("d = %f\n", d);
	if (isnan(d))
	{
		printf("d is NAN\n");
	}
	else
	{
		printf("d is not NAN\n");
	}

	d = 0.05;
	printf("d = %f\n", d);
	if (isnan(d))
	{
		printf("d is NAN\n");
	}
	else
	{
		printf("d is not NAN\n");
	}

输出结果：

d = nan
d is NAN
d = nan
d is NAN
d = 0.050000
d is not NAN