|
Armbruster, Springer, 1987}
Introduction to MACSYMA1
Richard H. Hand Dept. of Theoretical and Applied Mechanics, Cornell University2
Jul 10, 2000
Contents 1 Introduction
2 Special keys and symbols
3 Arithmetic
4 Algebra
5 Calculus
6 Matrix calculations
7 Programming in MAXIMA
8 A partial list of MAXIMA functions
1 Introduction
To invoke MAXIMA in Linux, type
maxima<ret>
The computer will display a greeting of the sort:
GCL (GNU Common Lisp) Version(2.3) ter jun 27 14:16:29 BRT 2000
Licensed under GNU Library General Public License
Contains Enhancements by W. Schelter
Maxima 5.4 ter jun 27 14:16:11 BRT 2000 (with enhancements by W. Schelter).
Licensed under the GNU Public License (see file COPYING)
(C1)
The (C1) is a ``label''. Each input or output line is labelled and can be referred to by its
own label for the rest of the session. C labels denote your commands and D labels
denote Displays of the machine's response. Never use variable names like C1 or
D5, as these will be confused with the lines so labeled.
MAXIMA is pragmatic about lower and upper case: regardless of your typing sin(x) or
SIN(x),
%e^x or %E^x, it will understand that you mean the sine and exponential
functions, and will echo in standard uppercase SIN and %E.
That doesn't apply to user variables, though: x and X are different
variables for MAXIMA ! (Try it.)
2 Special keys and symbols
- To end a MAXIMA session, type quit();. If you type ^C, here is what happens:
Correctable error: Console interrupt.
Signalled by MACSYMA-TOP-LEVEL.
If continued: Type :r to resume execution, or :q to quit to top level.
Broken at SYSTEM:TERMINAL-INTERRUPT. Type :H for Help.
MAXIMA>>:q
(C1)
Notice that typing :q or :t (for top level) after the MAXIMA>> prompt gets
you back to the MAXIMA level.
^Y, on the other hand, won't have any effect but being echoed on the screen; finally
^Z will have the same effect as quit();. (Here ^ stands for the control key, so
that ^C means first press the key marked control and hold it down while pressing the C key.)
- To abort a computation without leaving MAXIMA, type ^C. It is important for you to
know how to do this in case, for example, you begin a computation which is taking too long.
Remember to type :q at the MAXIMA>> prompt to return to MAXIMA. For example:
(C1) sum(1/x^2,x,1,1000);
Correctable error: Console interrupt.
Signalled by MACSYMA-TOP-LEVEL.
If continued: Type :r to resume execution, or :q to quit to top level.
Broken at SYSTEM:TERMINAL-INTERRUPT. Type :H for Help.
MAXIMA>>:q
(C2)
- In order to tell MAXIMA that you have finished your command, use the semicolon (;),
followed by a return. Note that the return key alone does not signal that you are done with your
input.
- An alternative input terminator to the semicolon (;) is the dollar sign ($),
which, however, supresses the display of MAXIMA's computation. This is useful if you are computing
some long intermediate result, and you don't want to waste time having it displayed on the screen.
- If you wish to repeat a command which you have already given, say on line (C5), you may
do so without typing it over again by preceding its label with two single quotes (''), i.e.,
''C5. (Note that simply inputing C5 will not do the job - try it.)
- If you want to refer to the immediately preceding result computed my MAXIMA, you can either
use its D label, or you can use the special symbol percent (%).
- The standard quantities e (natural log base), i (square root of -1) and p
(3.14159¼) are respectively referred to as %e (or %E), %i
(or %I), and %pi (or %PI). Note that the use of % here as a prefix
is completely unrelated to the use of % to refer to the preceding result computed.
- In order to assign a value to a variable, MAXIMA uses the colon (:), not the equal
sign. The equal sign is used for representing equations.
3 Arithmetic
The common arithmetic operations are
- [+] addition
- [-] subtraction
- [*] scalar multiplication
- [/] division
- [^] or ** exponentiation
- [.] matrix multiplication
- [sqrt(x)] square root of x.
MAXIMA's output is characterized by exact (rational) arithmetic. E.g.,
(C1) 1/100+1/101;
201
(D1) -----
10100
If irrational numbers are involved in a computation, they are kept in symbolic form:
(C2) (1+sqrt(2))^5;
5
(D2) (SQRT(2) + 1)
(C3) expand(%);
(D3) 29 SQRT(2) + 41
However, it is often useful to express a result in decimal notation. This may be accomplished by
following the expression you want expanded by ``,numer'':
(C4) %,numer;
(D4) 82.01219330881976
Note the use here of %
to refer to the previous result. In this version of MAXIMA, numer gives 16 significant
figures, of which the last is often unreliable. However, MAXIMA can offer arbitrarily high
precision by using the bfloat function:
(C5) bfloat(d3);
(D5) 8.201219330881976B1
The number of significant figures displayed is controlled by the MAXIMA variable FPPREC, which
has the default value of 16:
(C6) fpprec;
(D6) 16
Here we reset FPPREC to yield 100 digits:
(C7) fpprec:100;
(D7) 100
(C8) ''c5;
(D8) 8.20121933088197564152489730020812442785204843859314941221237124017312418#
7540110412666123849550160561B1
Note the use of two single quotes ('') in (C8) to repeat command (C5). MAXIMA can
handle very large numbers without approximation:
(C9) 100!;
(D9) 9332621544394415268169923885626670049071596826438162146859296389521759999#
322991560894146397615651828625369792082722375825118521091686400000000000000000#
0000000
4 Algebra
MAXIMA's importance as a computer tool to facilitate analytical calculations becomes more evident
when we see how easily it does algebra for us. Here's an example in which a polynomial is expanded:
(C1) (x+3*y+x^2*y)^3;
2 3
(D1) (x y + 3 y + x)
(C2) expand(%);
6 3 4 3 2 3 3 5 2 3 2 2 4
(D2) x y + 9 x y + 27 x y + 27 y + 3 x y + 18 x y + 27 x y + 3 x y
2 3
+ 9 x y + x
Now suppose we wanted to substitute 5/z for x in the above expression:
(C3) d2,x=5/z;
2 3 2 3 2
135 y 675 y 225 y 2250 y 125 5625 y 1875 y 9375 y
(D3) ------ + ------ + ----- + ------- + --- + ------- + ------ + -------
z 2 2 3 3 4 4 5
z z z z z z z
3
15625 y 3
+ -------- + 27 y
6
z
|
The MAXIMA function RATSIMP will place this over a common denominator:
(C4) ratsimp(%);
3 6 2 5 3 4 2 3
(D4) (27 y z + 135 y z + (675 y + 225 y) z + (2250 y + 125) z
3 2 2 3 6
+ (5625 y + 1875 y) z + 9375 y z + 15625 y )/z
Expressions may also be factored:
(C5) factor(%);
2 3
(3 y z + 5 z + 25 y)
(D5) ----------------------
6
z
MAXIMA can obtain exact solutions to systems of nonlinear algebraic equations. In this example we
solve three equations in the three unknowns A, B, C:
(C6) a + b*c=1;
(D6) b C + a = 1
(C7) b - a*c=0;
(D7) b - a C = 0
(C8) a+b=5;
(D8) b + a = 5
(C9) solve([d6,d7,d8],[a,b,c]);
25 SQRT(79) %I + 25 5 SQRT(79) %I + 5 SQRT(79) %I + 1
(D9) [[a = -------------------, b = -----------------, C = ---------------],
6 SQRT(79) %I - 34 SQRT(79) %I + 11 10
25 SQRT(79) %I - 25 5 SQRT(79) %I - 5 SQRT(79) %I - 1
[a = -------------------, b = -----------------, C = - ---------------]]
6 SQRT(79) %I + 34 SQRT(79) %I - 11 10
Note that the display consists of a ``list'', i.e., some expression contained between two brackets
[ ... ], which itself contains two lists. Each of the latter contain a distinct solution
to the simultaneous equations.
Trigonometric identities are easy to manipulate in MAXIMA. The function trigexpand uses the
sum-of-angles formulas to make the argument inside each trig function as simple as possible:
(C10) sin(u+v)*cos(u)^3;
3
(D10) COS (u) SIN(v + u)
(C11) trigexpand(%);
3
(D11) COS (u) (COS(u) SIN(v) + SIN(u) COS(v))
The function trigreduce, on the other hand, converts an expression into a form which is a sum
of terms, each of which contains only a single sin or cos:
(C12) trigreduce(d10);
SIN(v + 4 u) + SIN(v - 2 u) 3 SIN(v + 2 u) + 3 SIN(v)
(D12) --------------------------- + -------------------------
8 8
The functions realpart and imagpart will return the real and imaginary parts of a
complex expression:
(C13) w:3+k*%i;
(D13) %I k + 3
(C14) w^2*%e^w;
2 %I k + 3
(D14) (%I k + 3) %E
(C15) realpart(%);
3 2 3
(D15) %E (9 - k ) COS(k) - 6 %E k SIN(k)
5 Calculus
MAXIMA can compute derivatives and integrals, expand in Taylor series, take limits, and obtain exact
solutions to ordinary differential equations. We begin by defining the symbol f to be the
following function of x:
(C1) f:x^3*%E^(k*x)*sin(w*x);
3 k x
(D1) x %E SIN(w x)
We compute the derivative of f with respect to x:
(C2) diff(f,x);
3 k x 2 k x 3 k x
(D2) k x %E SIN(w x) + 3 x %E SIN(w x) + w x %E COS(w x)
Now we find the indefinite integral of f with respect to x:
(C3) integrate(f,x);
6 3 4 5 2 7 3
(D3) (((k w + 3 k w + 3 k w + k ) x
6 2 4 4 2 6 2 4 3 2 5
+ (3 w + 3 k w - 3 k w - 3 k ) x + (- 18 k w - 12 k w + 6 k ) x
4 2 2 4 k x
- 6 w + 36 k w - 6 k ) %E SIN(w x)
7 2 5 4 3 6 3 5 3 3 5 2
+ ((- w - 3 k w - 3 k w - k w) x + (6 k w + 12 k w + 6 k w) x
5 2 3 4 3 3 k x
+ (6 w - 12 k w - 18 k w) x - 24 k w + 24 k w) %E COS(w x))
8 2 6 4 4 6 2 8
/(w + 4 k w + 6 k w + 4 k w + k )
A slight change in syntax gives definite integrals:
(C4) integrate(1/x^2,x,1,inf);
(D4) 1
(C5) integrate(1/x,x,0,inf);
Integral is divergent
-- an error. Quitting. To debug this try DEBUGMODE(TRUE);)
Next we define the simbol g in terms of f (previously defined in C1) and the
hyperbolic sine function, and find its Taylor series expansion (up to, say, order 3 terms) about the
point x = 0:
(C6) g:f/sinh(k*x)^4;
3 k x
x %E SIN(w x)
(D6) -----------------
4
SINH (k x)
(C7) taylor(g,x,0,3);
2 3 2 2 3 3
w w x (w k + w ) x (3 w k + w ) x
(D7)/T/ -- + --- - -------------- - ---------------- + . . .
4 3 4 3
k k 6 k 6 k
|
The limit of g as x goes to 0 is computed as follows:
(C8) limit(g,x,0);
w
(D8) --
4
k
MAXIMA also permits derivatives to be represented in unevaluated form (note the quote):
(C9) 'diff(y,x);
dy
(D9) --
dx
The quote operator in (C9) means ``do not evaluate''. Without it, MAXIMA would have obtained
0:
(C10) diff(y,x);
(D10) 0
Using the quote operator we can write differential equations:
(C11) 'diff(y,x,2) + 'diff(y,x) + y;
2
d y dy
(D11) --- + -- + y
2 dx
dx
MAXIMA's ODE2 function can solve first and second order ODE's:
(C12) ode2(d11,y,x);
- x/2 SQRT(3) x SQRT(3) x
(D12) y = %E (%K1 SIN(---------) + %K2 COS(---------))
2 2
6 Matrix calculations
MAXIMA can compute the determinant, inverse and eigenvalues and eigenvectors of matrices which have
symbolic elements (i.e., elements which involve algebraic variables.) We begin by entering a matrix
m element by element:
(C1) m:entermatrix(3,3);
Is the matrix 1. Diagonal 2. Symmetric 3. Antisymmetric 4. General
Answer 1, 2, 3 or 4 : 4;
Row 1 Column 1: 0;
Row 1 Column 2: 1;
Row 1 Column 3: a;
Row 2 Column 1: 1;
Row 2 Column 2: 0;
Row 2 Column 3: 1;
Row 3 Column 1: 1;
Row 3 Column 2: 1;
Row 3 Column 3: 0;
Matrix entered.
[ 0 1 a ]
[ ]
(D1) [ 1 0 1 ]
[ ]
[ 1 1 0 ]
Next we find its transpose, determinant and inverse:
(C2) transpose(m);
[ 0 1 1 ]
[ ]
(D2) [ 1 0 1 ]
[ ]
[ a 1 0 ]
(C3) determinant(m);
(D3) a + 1
(C4) invert(m),detout;
[ - 1 a 1 ]
[ ]
[ 1 - a a ]
[ ]
[ 1 1 - 1 ]
(D4) -----------------
a + 1
In (C4), the modifier DETOUT keeps the determinant outside the inverse. As a check, we
multiply m by its inverse (note the use of the period to represent matrix multiplication):
(C5) m.d4;
[ - 1 a 1 ]
[ ]
[ 1 - a a ]
[ 0 1 a ] [ ]
[ ] [ 1 1 - 1 ]
(D5) [ 1 0 1 ] . -----------------
[ ] a + 1
[ 1 1 0 ]
(C6) expand(%);
[ a 1 ]
[ ----- + ----- 0 0 ]
[ a + 1 a + 1 ]
[ ]
[ a 1 ]
(D6) [ 0 ----- + ----- 0 ]
[ a + 1 a + 1 ]
[ ]
[ a 1 ]
[ 0 0 ----- + ----- ]
[ a + 1 a + 1 ]
(C7) factor(%);
[ 1 0 0 ]
[ ]
(D7) [ 0 1 0 ]
[ ]
[ 0 0 1 ]
In order to find the eigenvalues and eigenvectors of m, we use the function
EIGENVECTORS:
(C8) eigenvectors(m);
Warning - you are redefining the MACSYMA function EIGENVALUES
Warning - you are redefining the MACSYMA function EIGENVECTORS
SQRT(4 a + 5) - 1 SQRT(4 a + 5) + 1
(D8) [[[- -----------------, -----------------, - 1], [1, 1, 1]],
2 2
SQRT(4 a + 5) - 1 SQRT(4 a + 5) - 1
[1, - -----------------, - -----------------],
2 a + 2 2 a + 2
SQRT(4 a + 5) + 1 SQRT(4 a + 5) + 1
[1, -----------------, -----------------], [1, - 1, 0]]
2 a + 2 2 a + 2
In D8, the first triple gives the eigenvalues of m and the next gives their respective
multiplicities (here each is unrepeated). The next three triples give the corresponding
eigenvectors of m. In order to extract from this expression one of these eigenvectors, we may
use the PART function:
(C9) part(%,2);
SQRT(4 a + 5) - 1 SQRT(4 a + 5) - 1
(D9) [1, - -----------------, - -----------------]
2 a + 2 2 a + 2
|
7 Programming in MAXIMA
So far, we have used MAXIMA in the interactive mode, rather like a calculator. However, for
computations which involve a repetitive sequence of commands, it is better to execute a program.
Here we present a short sample program to calculate the critical points of a function f of two
variables x and y. The program cues the user to enter the function f, then it
computes the partial derivatives fx and fy, and then it uses the MAXIMA
command SOLVE to obtain solutions to fx = fy = 0. The program is written outside
of MAXIMA with a text editor, and then loaded into MAXIMA with the BATCH command. Here is the
program listing:
/* --------------------------------------------------------------------------
this is file critpts.max:
as you can see, comments in maxima are like comments in C
Nelson Luis Dias, nldias@simepar.br
created 20000707
updated 20000707
--------------------------------------------------------------------------- */
critpts():=(
print("program to find critical points"),
/* ---------------------------------------------------------------------------
asks for a function
--------------------------------------------------------------------------- */
f:read("enter f(x,y)"),
/* ---------------------------------------------------------------------------
echoes it, to make sure
--------------------------------------------------------------------------- */
print("f = ",f),
/* ---------------------------------------------------------------------------
produces a list with the two partial derivatives of f
--------------------------------------------------------------------------- */
eqs:[diff(f,x),diff(f,y)],
/* ---------------------------------------------------------------------------
produces a list of unknowns
--------------------------------------------------------------------------- */
unk:[x,y],
/* ---------------------------------------------------------------------------
solves the system
--------------------------------------------------------------------------- */
solve(eqs,unk)
)$
The program (which is actually a function with no argument) is called critpts. Each line is a
valid MAXIMA command which could be executed from the keyboard, and which is separated by the next
command by a comma. The partial derivatives are stored in a variable named eqs, and the
unknowns are stored in unk. Here is a sample run:
(C1) batch("critpts.max");
batching #/home/nldias/work/papers2000/intromax/critpts.max
(C2)
(C2) critpts() := (PRINT("program to find critical points"),
f : READ("enter f(x,y)"), PRINT("f = ", f), eqs : [DIFF(f, x), DIFF(f, y)],
unk : [x, y], SOLVE(eqs, unk))
(C3) critpts();
program to find critical points
enter f(x,y)
%e^(x^3+y^2)*(x+y);
2 3
y + x
f = (y + x) %E
(D3) [[x = 0.4588955685487 %I + 0.35897908710869,
y = 0.49420173682751 %I - 0.12257873677837],
[x = 0.35897908710869 - 0.4588955685487 %I,
y = - 0.49420173682751 %I - 0.12257873677837],
[x = 0.41875423272348 %I - 0.69231242044203,
y = 0.4559120701117 - 0.86972626928141 %I],
[x = - 0.41875423272348 %I - 0.69231242044203,
y = 0.86972626928141 %I + 0.4559120701117]]
8 A partial list of MAXIMA functions
See the MAXIMA Manual in the maxima-5.4/info/ directory in texinfo or html format. From
MAXIMA itself, you can use DESCRIBE(function name).
- ALLROOTS(A)
- Finds all the (generally complex) roots of the polynomial equation
A, and lists them in NUMERical format (i.e. to 16 significant figures).
- APPEND(A,B)
- Appends the list B to the list A, resulting in a single
list.
- BATCH(A)
- Loads and runs a BATCH program with filename A.
- COEFF(A,B,C)
- Gives the coefficient of B raised to the power C in
expression A.
- CONCAT(A,B)
- Creates the symbol AB.
- CONS(A,B)
- Adds A to the list B as its first element.
- DEMOIVRE(A)
- Transforms all complex exponentials in A to their trigonometric
equivalents.
- DENOM(A)
- Gives the denominator of A.
- DEPENDS(A,B)
- Declares A to be a function of B. This is useful for
writing unevaluated derivatives, as in specifying differential equations.
- DESOLVE(A,B)
- Attempts to solve a linear system A of ODE's for unknowns B
using Laplace transforms.
- DETERMINANT(A)
- Returns the determinant of the square matrix A.
- DIFF(A,B1,C1,B2,C2,...,Bn,Cn)
- Gives the mixed partial derivative of A with
respect to each Bi, Ci times. For brevity, DIFF(A,B,1) may be represented by
DIFF(A,B). 'DIFF(...) represents the unevaluated derivative, useful in specifying
a differential equation.
- EIGENVALUES(A)
- Returns two lists, the first being the eigenvalues of the square
matrix A, and the second being their respective multiplicities.
- EIGENVECTORS(A)
- Does everything that EIGENVALUES does, and adds a list of the
eigenvectors of A.
- ENTERMATRIX(A,B)
- Cues the user to enter an A × B matrix,
element by element.
- EV(A,B1,B2,...,Bn)
- Evaluates A subject to the conditions Bi. In
particular the Bi may be equations, lists of equations (such as that returned by
SOLVE), or assignments, in which cases EV ``plugs'' the Bi into A. The
Bi may also be words such as NUMER (in which case the result is returned in numerical
format), DETOUT (in which case any matrix inverses in A are performed with the
determinant factored out), or DIFF (in which case all differentiations in A are
evaluated, i.e., 'DIFF in A is replaced by DIFF). For brevity in a manual
command (i.e., not inside a user-defined function), the EV may be dropped, shortening the
syntax to A,B1,B2,...,Bn.
- EXPAND(A)
- Algebraically expands A. In particular multiplication is
distributed over addition.
- EXPONENTIALIZE(A)
- Transforms all trigonometric functions in A to their complex
exponential equivalents.
- FACTOR(A)
- Factors A.
- FREEOF(A,B)
- Is true if the variable A is not part of the expression B.
- GRIND(A)
- Displays a variable or function A in a compact format. When used
with WRITEFILE and an editor outside of MAXIMA, it offers a scheme for producing
BATCH files which include MAXIMA-generated expressions.
- IDENT(A)
- Returns an A × A identity matrix.
- IMAGPART(A)
- Returns the imaginary part of A.
- INTEGRATE(A,B)
- Attempts to find the indefinite integral of A with respect to
B.
- INTEGRATE(A,B,C,D)
- Attempts to find the indefinite integral of A with respect to
B. taken from B = C to B = D. The limits of integration C and
D may be taken is INF (positive infinity) of MINF (negative infinity).
- INVERT(A)
- Computes the inverse of the square matrix A.
- KILL(A)
- Removes the variable A with all its assignments and properties from
the current MAXIMA environment.
- LIMIT(A,B,C)
- Gives the limit of expression A as variable B approaches
the value C. The latter may be taken as INF of MINF as in INTEGRATE.
- LHS(A)
- Gives the left-hand side of the equation A.
- LOADFILE(A)
- Loads a disk file with filename A from the current default
directory. The disk file must be in the proper format (i.e. created by a SAVE command).
- MAKELIST(A,B,C,D)
- Creates a list of A's (each of which presumably depends on
B), concatenated from B = C to B = D
- MAP(A,B)
- Maps the function A onto the subexpressions of B.
- MATRIX(A1,A2,...,An)
- Creates a matrix consisting of the rows Ai, where each
row Ai is a list of m elements, [B1, B2, ..., Bm].
- NUM(A)
- Gives the numerator of A.
- ODE2(A,B,C)
- Attempts to solve the first- or second-order ordinary differential
equation A for B as a function of C.
- PART(A,B1,B2,...,Bn)
- First takes the B1th part of A, then the
B2th part of that, and so on.
- PLAYBACK(A)
- Displays the last A (an integer) labels and their associated
expressions. If A is omitted, all lines are played back. See the Manual for other
options.
- RATSIMP(A)
- Simplifies A and returns a quotient of two polynomials.
- REALPART(A)
- Returns the real part of A.
- RHS(A)
- Gives the right-hand side of the equation A.
- SAVE(A,B1,B2,..., Bn)
- Creates a disk file with filename A in the current
default directory, of variables, functions, or arrays Bi. The format of the file permits
it to be reloaded into MAXIMA using the LOADFILE command. Everything (including labels)
may be SAVEd by taking B1 equal to ALL.
- SOLVE(A,B)
- Attempts to solve the algebraic equation A for the unknown B. A
list of solution equations is returned. For brevity, if A is an equation of the form
C = 0, it may be abbreviated simply by the expression C.
- STRING(A)
- Converts A to MACSYMA's linear notation (similar to FORTRAN's) just as if
it had been typed in and puts A into the
buffer for possible editing. The STRING'ed expression should not be used in a computation.
- STRINGOUT(A,B1,B2,...,Bn)
- Creates a disk file with filename A in the current
default directory, of variables (e.g. labels) Bi. The file is in a text format and is not
reloadable into MAXIMA. However the strungout expressions can be incorporated into a FORTRAN,
BASIC or C program with a minimum of editing.
- SUBST(A,B,C)
- Substitutes A for B in C.
- TAYLOR(A,B,C,D)
- Expands A in a Taylor series in B about B = C,
up to and including the term (B-C)D. MAXIMA also supports Taylor expansions in more
than one independent variable; see the Manual for details.
- TRANSPOSE(A)
- Gives the transpose of the matrix A.
- TRIGEXPAND(A)
- Is a trig simplification function which uses the sum-of-angles
formulas to simplify the arguments of individual SIN's or COS's. For example,
trigexpand(sin(x+y)) gives COS(x) SIN(y) + SIN(x) COS(y).
- TRIGREDUCE(A)
- Is a trig simplification function which uses trig identities to
convert products and powers of SIN and COS into a sum of terms, each of which
contains only a single SIN or COS. For example, trigreduce(sin(x)^2) gives
(1 - COS(2x))/2.
- TRIGSIMP(A)
- Is a trig simplification function which replaces TAN, SEC,
etc., by their SIN and COS equivalents. It also uses the identity
SIN()2 + COS()2 = 1.
Footnotes:
1 Adapted from ``Perturbation Methods, Bifurcation Theory and Computer Algebra'' by Rand and
Armbruster, Springer, 1987
2 Adapted to LATEX and HTML by Nelson L. Dias (nldias@simepar.br), SIMEPAR Technological Institute and Federal University of Paranį, Brazil
File translated from TEX by TTH, version 2.51. On 10 Jul 2000, 18:41.
|