Generating random numbers with a normal distribution

I have a dataset that looks, when charted, like this:

To me, this looks like a normal/gaussian distribution. The mean is 0.0017 and the standard deviation is 0.0571. I am trying to generate random numbers along the same distribution using numpy's np.random.normal function as follows:

a = numpy.random.normal(0.0017, 0.0571)

This creates results that look like this:

As you can see, the sloping of the distribution is much gentler, with a much smaller peak, and the outliers tail off much more quickly. Is there a way to adjust my standard distribution so that my randomly generated numbers match the originals more closely?