Plotting Surfaces with GnuPlot-Python??

[Prem Rachakonda]

Hi,
  I have a problem in plotting contours/surfaces. The problem
description is as follows:

x--->
y z1-1 z1-2 z1-3 z1-4 z1-5.......z1-5000
| z2-1 z2-2 z2-3 z2-4 ...
| z3-1 z3-2 z3-3 z3-4 ...
V ...
  ...
  .
  .
  z100-1.........................z100-5000


I have the above matrix z of size [100,5000], each value representing
the height in z-axis(5000 points in x multiplied by 100 scans in y)

I have three matrices now, x[5000], and y[100] and z[100,5000]

And I would like to plot the contour(surface) with this x,y,z data. 

My python code goes like this:(Only a part of it is displayed here)

***************************************************
scalex=1.2
scaley=2.4
scalez=3.6
size=100
g=Gnuplot.Gnuplot()
g.clear()
g.("set view 90,0")

#The function 'dlmread' reads the matrix from a number of files(set by
'size' and the z will have the number of rows=size and the number of
columns='5000'

z=reshape((array(iotools.dlmread('Folder_1','data.txt',size,100,5100))*scalez),(5000,size))

x=(arange(0.,7800))*scalex

y=(arange(0,size))*scaley

g('set parametric')
g('set data style dots')
g('set contour base')
g.splot(Gnuplot.GridData(z,x,y, binary=0))

***************************************************

If is set 'size=1', ie. try to plot only one set of values to get the
profile of the surface, I get it absolutely right, but only a
profile..not the surface.

If I set size=2, ie. try to plot two sets of values, I get 4 plots of
the profile and the every value in x is halved.

If I set size=3, then I get 9 plots and x is reduced to 1/3rd

I am not sure I understood how 'splot' plots the surface. 

Please let me know if there is a way to solve this problem. Why does
the data gets scaled down???