how to get adress position inside vector

hi!

please for some help

i have this vector

i need to some value of eld in a vector gravlo just in right position given in vector adress

but matlab say Subscript indices must either be real positive integers or logicals.

can someone help me how to fix it

gravlo=zeros(1,100);

eld =[ 0

-1.7222

0

7.0000

0

-1.7778

0

6.8889

0

-1.7778

0

7.0000

0

-1.7222

0

7.1111];

adress =[ 0

3

0

2

0

1

31

32

61

62

63

64

65

66

33

34 ];

gravlo(adress)=gravlo(adress)+eld.*20;

[Arif Hoq]

can you show us your expected result ?

i need to some value of eld in a vector gravlo just in right position given in vector adress

which value of 'eld' do you want in 'gravlo' vector?

[alize beemiel]

thank for answering

but perhaps i dont explaine my self correctly

i have a big matrix over 1000 or more size (16,1000) name G witch containe adress of free element, the value zero is mean that the element dont need update or not free

and i have a other matrix gravlo who make sum for only the element need to be update according of adress G with the value of eld+somthing (something =20) for exemple

i use a loop

let's use loop for just 5 first element

like this

gravlo=zeros(1,100);

eld =[ 0

-1.7222

0

7.0000

0

-1.7778

0

6.8889

0

-1.7778

0

7.0000

0

-1.7222

0

7.1111];

G=[3 66 156 246 336

0 63 153 243 333

2 64 154 244 334

0 61 151 241 331

1 62 152 242 332

31 121 211 301 391

32 122 212 302 392

61 151 241 331 421

62 152 242 332 422

63 153 243 333 423

64 154 244 334 424

65 155 245 335 425

66 156 246 336 426

33 123 213 303 393

34 124 214 304 394];

for i=1:5

g=G(:,i)

gravlo(g)=gravlo(g)-eld*20;

end

%

%

i try this to axtract a non zero value this ... gravlo(g(g~=0))

but when a sum

the dimension is of course different

and i m not sure that the value is just in the corect adress according to g

for i=1:5

g=G(:,i)

gravlo(g(g~=0))

gravlo(g(g~=0))=gravlo(g(g~=0))-eld*20;

end

sorry for my bad explained

i m so confuse

[Torsten]

At least, eld and G should have the same number of rows.

This is not true in your case.

Even if they had the same number of rows, if you remove indices from g as you do, the line

gravlo(g(g~=0))=gravlo(g(g~=0))-eld*20;

will no longer work.

Further, "gravlo" must be a column, not a row vector.

[Arif Hoq]

You have these array (gravlo,eld,address,G). still i am not clear about your expectation.can you show some result ? like 4/5 rows of your expected result

gravlo=zeros(1,100);
eld =[0;-1.7222;0;7.0000;0; -1.7778;0;6.8889;0; -1.7778;0;7.0000;0;-1.7222;0;7.1111];
adress =[0;3;0;2;0;1;31;32;61;62;63;64;65;66;33;34 ];
G=[3	66	156	246	336
0	63	153	243	333
2	64	154	244	334
0	61	151	241	331
1	62	152	242	332
31	121	211	301	391
32	122	212	302	392
61	151	241	331	421
62	152	242	332	422
63	153	243	333	423
64	154	244	334	424
65	155	245	335	425
66	156	246	336	426
33	123	213	303	393
34	124	214	304	394];

[alize beemiel]

thank you Sir

i try this

i know its look like a sharabian!!!

but its work when i use my big matix with more than size=2120

so thank You for answering

clc
clear all


eld =[   0
    -1.7222
    0
    7.0000
    0
    -1.7778
    0
    6.8889
    0
    -1.7778
    0
    7.0000
    0
    -1.7222
    0
    7.1111];

G=[ 0       65   155   245  235           
    3	        66	156	246	336
    0	        63	153	243	333
    2	        64	154	244	334
    0	        61	151	241	331
    1	        62	152	242	332
    31	121	211	301	391
    32	122	212	302	392
    61	151	241	331	421
    62	152	242	332	422
    63	153	243	333	423
    64	154	244	334	424
    65	155	245	335	425
    66	156	246	336	426
    33	123	213	303	393
    34	124	214	304	394];

gravlo=zeros(427,1);

for  iel=1:5
    g=G(:,iel)
    ss=find(g)
    g(g~=0)
    [mm nn]=size(g(g~=0));
    [ee qq]=size(eld);
    if mm<ee
        edd=eld(ss);
        gravlo(g(g~=0))=gravlo(g(g~=0))-edd*20;% 
    else
        gravlo(g)= gravlo(g)-eld*20;%
    end
end

[Jan]

A simplified version of your code:

gravlo = zeros(427,1);
for iel=1:5
    g = G(:,iel);
    m = (g ~= 0);
    gravlo(g(m)) = gravlo(g(m)) - eld(m) * 20;
end

Hints: Do not use "clear all" in productive code, because it is a waste of time only.

Use the buttons on top of the field for posting messages to format the code. This improves the readability.

[alize beemiel]

oh !!!

thank you Sir,

its very nice and its work with a simply code

thanks very much Sir,

Answer 1

[Jan]

Copied from my comment as an answer:

gravlo = zeros(427,1);
for iel = 1:5
    g = G(:, iel);
    m = (g ~= 0);
    gravlo(g(m)) = gravlo(g(m)) - eld(m) * 20;
end