A coworker is trying to load a 1Gb text data file into a numpy array using numpy.loadtxt, but he says it is using up all of his machine's 6Gb of RAM. Is there a more efficient way to read such text data files? -- Russell