DMelt:IO/Saving Arrays

From jWork.ORG
Jump to: navigation, search
Limitted access. Reguest membership or login to this link first if you are already a member

Multi-dimensional arrays

Arrays in 1,2,3,4,5 dimensions can be written in a binary form to files and later restored. Unlike the standard Java serialization, we will consider a binary form for such arrays.

Input and Output for 1D and 2D array

The easiest way to store 1D arrays of double values in a compact binary form is to use jhplot.math.DoubleArray or jhplot.math.IntegerArray (for integers) and then use the to write and read such arrays.

Here is a small example:

from  jhplot.math.DoubleArray import *
from import *

a =[1.0, 2.0, 3.0, 4.0]
b =[ [1.0, 2.0, 3.0, 4.0], [11.0, 12.0, 13.0, 14.0] ]
writeDoubleArray(File("a.txt"), a) # write files in ASCII
writeDoubleArray(File("b.txt"), b)

a_read = readDouble1DArray(File("a.txt")) # reading arrays
b_read = readDoubleArray(File("b.txt"))

print "a_read =  \n" + printDoubleArray(a_read)  # now print
print "b_read =  \n" + printDoubleArray(b_read)

Large arrays can be written in the binary form:

from  jhplot.math.DoubleArray import *
from import *
a =[1.0, 2.0, 3.0, 4.0]
writeDoubleArray(File("a.dat"), a, "LITTLE_ENDIAN")      # write the array  
a_read = readDoubleArray(File("a.dat"), "LITTLE_ENDIAN") # read these file
print  "a_read from binary file = \n" + printDoubleArray(a_read)

2D arrays cannot be written in the binary form using this approach

Below we will consider a more general approach to write and read 1D, 2D, 3D, 4D and 5D arrays in the binary form.

I/O for 1D array

I/O for 2D array

I/O of 3D array

I/O of 4D array

I/O of 5D array

More information on this topic is in DMelt books