 Not returning correct Matrix Calculation - Printable Version +- HP Forums (https://www.hpmuseum.org/forum) +-- Forum: HP Calculators (and very old HP Computers) (/forum-3.html) +--- Forum: HP Prime (/forum-5.html) +--- Thread: Not returning correct Matrix Calculation (/thread-16823.html) Not returning correct Matrix Calculation - Amer7 - 04-27-2021 06:47 PM My calculator is not giving me the correct matrix calculation. Code: ``` 2568.0000    -2568.00            1    0 2568.0000    2568.00000    0    1 3852.1284    -2568.00            1    0 2568.0000    3852.12840    0    1 3536.9834    -3411.36           1    0 3411.3569    3536.98344    0    1 That is Matrix A Matrix O  -4851346.102 -6513502.217 -4850456.274 -6513809.817 -4850472.619 -6513149.860 TRN(A)*A should be 58770372.8276     0.0000              9957.1118    8547.3569 0.0000                   58770372.8276     -8547.3569    9957.1118 9957.1118           -8547.3569      3.0000            0.0000 8547.3569            9957.1118      0.0000            3.0000 Inverse(TRN(A)*A) should be 7.29974E-07    -1.81352E-22    -0.002422813    -0.002079784 -2.53893E-21    7.29974E-07    0.002079784    -0.002422813 -0.002422813    0.002079784    14.30029098    0 -0.002079784    -0.002422813    -4.54077E-14    14.30029098 But I get different numbers than in Excel, im using Shift + devide key to inverse. Can someone explain, and later in my calculation i get wrong answer because of Inverse mistake. TRN(A)*O should be -1.03972E+11 -23394687040 -1.46E+07 -1.95E+07 - I get this OK, But my issue is here M=( Inverse((TRN(A)*A) * (TRN(A)*O) -0.6929740 0.2395352 4853740.802 6514666.637 But I get: 0.6929740 -0.2395352 -4853740.802 -6514666.637 Can someone explain why this is happening? - Im not writing a program im manually doing the calculation on Home screen. And later  A*M+O should be 0.017 -0.011 -0.023 -0.017 0.006 0.027 but I get 0.03314 -0.01158  etc. - WRONG I have tried to multiply the M matrix with (-1) but again calculation of O seems wrong.``` What is the best command for finding Inverse matrix ? same happens when i use inv(Matrix that i want to invert) [attachment=9421] RE: Not returning correct Matrix Calculation - Albert Chan - 04-27-2021 08:09 PM Hi, Amer7 Tried this in Numpy, Excel numbers were wrong. >>> from numpy import * >>> A = array([ ... [2568.0000, -2568.00 , 1, 0], ... [2568.0000, 2568.00000, 0, 1], ... [3852.1284, -2568.00 , 1, 0], ... [2568.0000, 3852.12840, 0, 1], ... [3536.9834, -3411.36 , 1, 0], ... [3411.3569, 3536.98344, 0, 1]]) >>> B = dot(transpose(A), A) Code: ```>>> print B [[  5.87703727e+07  -1.08281943e+01   9.95711180e+03   8.54735690e+03]  [ -1.08281943e+01   5.87703941e+07  -8.54736000e+03   9.95711184e+03]  [  9.95711180e+03  -8.54736000e+03   3.00000000e+00   0.00000000e+00]  [  8.54735690e+03   9.95711184e+03   0.00000000e+00   3.00000000e+00]] >>> print linalg.inv(B) [[  7.29974488e-07   3.48034952e-13  -2.42281154e-03  -2.07978531e-03]  [  3.48034952e-13   7.29972621e-07   2.07977844e-03  -2.42280733e-03]  [ -2.42281154e-03   2.07977844e-03   1.43002735e+01   1.61304069e-05]  [ -2.07978531e-03  -2.42280733e-03   1.61304069e-05   1.43002770e+01]]``` RE: Not returning correct Matrix Calculation - Amer7 - 04-27-2021 08:20 PM Thank you Albert for numerous replies, I get the same numbers as you on my calculator. I guess something is wrong in Excel ( all formulas seemed ok). - I really doubt these are calculations from the teacher. The excel all matches up, with control calculations. Even my friends that use TI get the same numbers RE: Not returning correct Matrix Calculation - Albert Chan - 04-27-2021 10:56 PM (04-27-2021 08:20 PM)Amer7 Wrote:  I guess something is wrong in Excel ( all formulas seemed ok). It is possible the problem is not Excel. From your attached screenshot, there were a picture suggesting A51 = A62, A61 = -A52. Your provided A (row 5 and row 6) does not match this symmetrical pattern. Not terribly off, but enough to mess up the numbers. RE: Not returning correct Matrix Calculation - Amer7 - 04-28-2021 05:25 AM (04-27-2021 10:56 PM)Albert Chan Wrote:   (04-27-2021 08:20 PM)Amer7 Wrote:  I guess something is wrong in Excel ( all formulas seemed ok). It is possible the problem is not Excel. From your attached screenshot, there were a picture suggesting A51 = A62, A61 = -A52. Your provided A (row 5 and row 6) does not match this symmetrical pattern. Not terribly off, but enough to mess up the numbers. I have found the Issue, Im not sure why this happens. When number in column F have in front of them - ( minus, aka they are negative) Inverse matrix is not going to be the same as in Excel. So when I input O matrix as all positive numbers I get inverse matrix correct. And in last matrix calculation ( excel uses formula example (trn(A)*N) + O and when I adjust this to (trn(A)*N) - O I get everything correct. I'm not sure why this difference is happening between calculator and Excel. -Downloaded a virtual TI Nspire to check, they also get the same Matrix as we on our HP Prime. -So its something querk of Excel.