I have a loop in which I generate approximately 100 mp matrices with sizes ranging form N = 4 to ~1500. In order to use them later I save them as cell entries (A{i} = A). In some instances these will save, other times I get a corrupt file error warning.

It always happens with the same matrices, just wth some parameters I get the errors with others I don't.

Any ideas or suggestions?

Wonder if it is possible, currently, to compile your libraries using MCC for eventual deployment as a standalone application.

I've noticed that when using advanpix data structures with eigs(), I get erroneous results when requesting the smallest eigenvalue. 

Please refer to the following code using Matlab's data structure:

A_matlab = speye(25,25); %initialize Matlab sparse identity matrix
A_matlab(13,13) = 1/3;
eigs(A_matlab,1,'sa') %request smallest eigenvalue

When running this code, I receive the following output:

ans =

   0.333333333333333

Now consider a similar experiment with Advanpix:

A_advanpix = mp(speye(25,25),32); %initialize Advanpix sparse identity matrix
A_advanpix(13,13) = mp('1/3');
eigs(A_advanpix,1,'sa') %request smallest eigenvalue

When running this code, I receive the following output:  

ans = 

    0

I have noticed if I replace '1/3' with '-1/3' in the above experiments, I obtain the correct results, namely both return "-1/3" in the appropriate precision. I appreciate any help with this issue, thank you for your assistance!

I've noticed when doing vectorized operations with advanpix mp data structures, it's extremely slow compared to Matlab. For example, consider the follow code where I initialize 1000x1000 sparse, random matrices using both Matlab double and advanpix double data structures:

A_matlab = sprand(1000,1000,.2);
A_advanpix = mp(sprand(1000,1000,.2),16);

I then time the operation of "zeroing out" the first 100 rows in each

tic
A_matlab(1:100,:) = 0;
toc
Elapsed time is 0.008518 seconds.

tic
A_advanpix(1:100,:) = 0;
toc
Elapsed time is 2.534518 seconds.

Of course, this naturally gets much worse as the size of A increases. Is this behavior expected or am I using the advanpix data structures incorrectly?

Thank you!

Joe

Is it possible to specify the number of bits used to represent numbers, rather than the number of decimal digits?

>> mp(1/3)            % Accuracy is limited to double precision
ans =                 % since 1/3 was evaluated in double precision first
    0.33333333333333331482961625624739099293947219848633

How are the extended digits, i.e. 1482961625624739099293947219848633, created? Are they totally random numbers or some of them are determined by the guard digits in the double-precision representation of 1/3?

I've been thinking that a natural extension would be done by trailing zeros, resulting, for example, in the example above, something like:

0.333333333333333300000000000000000000000000000000

Any discussion/comment is welcome. Thanks!

On R201b (Linux Mint 19.2) is the following problem:

>> mp.Test
mp.Digits() : <- success
mp.GuardDigits() : <- success
double() : <- success
int8() : <- success
uint8() : <- success
int16() : <- success
uint16() : <- success
int32() : <- success
uint32() : <- success
int64() : <- success
uint64() : <- success
colon() : <- success
plus() : <- success
minus() : <- success
times() : <- success
mtimes() : <- success
rdivide() : <- success
ldivide() : <- success
mldivide() : <- success
mrdivide() : <- success
mpower() : <- success
power() : <- success
realpow() : <- success
transpose() : <- success
ctranspose() : <- success
uminus() : <- success
uplus() : <- success
sin() : <- success
cos() : <- success
tan() : <- success
sec() : <- success
csc() : <- success
cot() : <- success
acos() : <- success
asin() : <- success
atan() : <- success
acot() : <- success
atan2() : <- success
hypot() : <- success
cosh() : <- success
sinh() : <- success
tanh() : <- success
sech() : <- success
csch() : <- success
coth() : <- success
acosh() : <- success
asinh() : <- success
atanh() : <- success
acoth() : <- success
asech() : <- success
acsch() : <- success
exp() : <- success
expm1() : <- success
log() : <- success
log10() : <- success
log1p() : <- success
log2() : <- success
nextpow2() : <- success
pow2() : <- success
sqrt() : <- success
reallog() : <- success
realsqrt() : <- success
nthroot() : <- success
pow2(F,E) : <- success
min(), max() : <- success
prod(matrix) : <- success
prod(vector) : <- success
sum(matrix) : <- success
sum(vector) : <- success
cumsum(matrix) : <- success
cumsum(vector) : <- success
cumprod(matrix) : <- success
cumprod(vector) : <- success
dot() : <- success
cross() : 2 <- fail
3 <- fail
9 <- fail
10 <- fail
<- success
svd() : <- success
qr() : <- success
lu(square) : <- success
lu(rect) : <- success
pinv() : <- success
null() : <- success
balance() : <- success
eig() : <- success
qz() : <- success
hess() : <- success
chol() : <- success
schur() : <- success
ordschur() : <- success
rank() : <- success
trace() : <- success
det() : <- success
inv() : <- success
sort(real) : <- success
sort(complex) : <- success
find() : <- success
<,<=,>,>=,==,~= : <- success
and,or,not,xor : <- success
all() : <- success
any() : <- success
isinf() : <- success
isnan() : <- success
isfinite() : <- success
isreal() : <- success
abs() : <- success
sign() : <- success
conj() : <- success
angle() : <- success
imag() : <- success
real() : <- success
complex() : <- success
ceil() : <- success
fix() : <- success
floor() : <- success
idivide() : <- success
round() : <- success
rem() : <- success
mod() : <- success
tril(matrix) : <- success
tril(vector) : <- success
triu(matrix) : <- success
triu(vector) : <- success
diag(matrix) : <- success
diag(vector) : <- success
norm(matrix) : <- success
norm(vector) : <- success
cond() : <- success
rcond() : <- success
factorial() : <- success
mean() : <- success
std() : <- success
erf() : <- success
erfc() : <- success
erfi() : <- success
FresnelS() : <- success
FresnelC() : <- success
gammaln() : <- success
gamma() : <- success
gammainc() : <- success
psi() : <- success
zeta() : <- success
eint() : <- success
logint() : <- success
cosint() : <- success
sinint() : <- success
besselj() : <- success
bessely() : <- success
besseli() : <- success
besselk() : <- success
besselh() : <- success
hypergeom() : <- success
KummerM() : <- success
KummerU() : <- success
expm() : <- success
logm() : <- success
sqrtm() : <- success
sinm() : <- success
cosm() : <- success
sinhm() : <- success
coshm() : <- success
fft : <- success
ifft : <- success
fft2 : <- success
ifft2 : <- success

Matlab creates an additional copy of that slice and since my array is huge, it takes a lot of time.
Is there any possibility to add some function like block operator in eigen3?

Like this one

https://ch.mathworks.com/matlabcentral/fileexchange/24576-inplacearray-a-semi-pointer-package-for-matlab

Looking for past versions of the multiprecision toolbox on the Advanpix website.

Where are they?

Now, precomputeLU only supports sparse matrix. In addition, Maybe also precomputeQR.