Data importing troubles.
Ed Murphy
I have scoured the web trying to find a ready-made solution for
converting the MDIF file to something that Matlab can import without
success. So, I am trying to figure out what I need to do. The data
was generated from a simulation controlled by 3 nested sweeps, so I
have data that is multidimensional and irregular, that is, the
dimensions are not necessarily the same shape. This leads me to
believe that I must somehow parse this data into multidimensional
cell arrays, since numeric matrices must be rectangular.
Here are some examples:
!==========================================================
! Example 1
REM This has 3 indepVars: v1, v2, v3(innermost) and
REM 4 depVars: dv1(integer), dv2(real), dv3(string) and
REM dv4(hexadecimal), but is read in as a string
! There are 2 data nodes, with an attribute "att3"
VAR v1(0) = 1
VAR v2(1) = 2.2
BEGIN blk1
# att3(2) = "Hello, World"
% v3(1) dv1(1) dv2(1) dv3(2) dv4(hexadecimal)
7.7 8 9.9999 "line 1" 0xabc
8.8 9 1.11 "line 2 " 0x123
END
VAR v1(0) = 2
VAR v2(1) = 3.2
BEGIN blk1
# att3(2) = "Hello, Universe "
% v3(1) dv1(1) dv2(1) dv3(2) dv4(hexadecimal)
8.7 9uF 10.9999mA "line 1" 0xff
9.8 10uF 11.11mA "line 2 " 0xdef
END
!==========================================================
! Example 2
! Created Tue Mar 9 13:39:19 1999
! Data Acquired Tue Mar 9 13:38:34 1999
BEGIN NDATA_noise
% freq(real) Sopt(complex) NFmin(real) Rn(real) PortZ[1](real)
1e+09 0.098481 0.017365 1 5 50
2e+09 0.18794 0.068404 2 10 50
3e+09 0.25981 0.15 3 15 50
4e+09 0.30642 0.25712 4 20 50
5e+09 0.32139 0.38302 5 25 50
6e+09 0.3 0.51962 6 30 50
7e+09 0.23941 0.65778 7 35 50
8e+09 0.13892 0.78785 8 40 50
9.543e+09 -0.014122 0.911 9.5445 46.166 50
END
This is just a generic example of the file format. If anyone knows
how to parse this into cell arrays, I would love to hear from you.
Assuming that no one wants to tackle this, I have an even more basic
question. For each dependent variable, I have either a real or a
complex value array, which is a function of perhaps 3 or 4
independent variables each with there own arrays of varying length.
How do I represent the data? Let me give a specific example.
VS_high = [0.8074;0.8043;0.8013;0.7985;0.7957;0.7930;...
0.9890;0.9840;0.9792;0.9745;0.9700;0.9656];
DC_Ibias = [0.005;0.01];
DC_Width = [0.0001;0.000102;0.000104;0.000106;0.000108;...
0.00011];
DC_freq = 0;
VS_high is a function of (DC_Ibias, DC_Width, DC_freq), so I know
that VS_high(1,3,1) = 0.8013; but how do I represent the fact that
the independent variables in this case are DC_Ibias = 0.005, DC_Width
= 0.000104, DC_freq = 0 ?
I have 4 values and I need to store them in a way that [d1,d2,d3] =
size(VS_high) reports that d1=2,d2=6,d3=1? Normally the the first
dim is the row value, the second dim is the column value and the
third dim is the page value, so I get the indexing I just can’t see
where the dep variable should be stored and where the independent
values should be stored. If anyone could clear this up for me I
would appreciate it.
Ed Murphy
I should add two more notes. First, I realize that above, I was a
little fast and loose with the order of the dimensions/indices.
Second, I see this data being stored in a lookup table configuration.
I am just not clear on how the inputs (independent variables) and
the outputs (dependent variables) should be structured.
tomcom
The description for the format of GMDIF can be found here:
http://cp.literature.agilent.com/litweb/pdf/ads2004a/cktsim/ck0418.html
and the description for X-parameter data in GMDIF format can be found here (scroll down a bit):
http://edocs.soco.agilent.com/display/ads2009U1/Working+with+Data+Files#WorkingwithDataFiles-XparameterGMDIFFormat
I'm attempting to write a parser but I'm having trouble finding an efficient way to do so. I think the main challenge is figuring out a good way to extract all of the independent variable values, especially since the header comment doesn't seem to be consistently generated between ADS and the NVNA software. Does anyone have any suggestions? Thank you!