Tips for Math with the TI-89 Calculator

© 2023 October 23, version 1.2, Dr. Wm J. Larson, International School of Geneva, [email protected], Corrections welcome.
 Graphing 5: ZoomSqr scales x & y the same, so circles
look round, squares look square and
To enter an equation to be graphed, key  Y=. lines look.
Type in the equation. 6: ZoomStd sets x & y min = -10, x & y max
Key F4  to select or deselect an equations. = 10 and x & y scl = 1. If nothing appears
Only selected equations are graphed. on the screen, try this first.
Key F6 Style to pick the way the graph is 7: ZoomTrig is useful for graphing trig
displayed (dotted, thick, etc.). This is useful functions. It sets the pixel size = /24 =
if more than one equation is selected. 7.5° and xscl = /2 = 90°.
If the window is too big (i.e. the graph is a tiny 9: ZoomData, for use with scatterplots or
unhelpful squiggle) or too little (i.e. the histograms data, sets xmin & xmax to
important features of the graph are off the match the data.
screen), try ZoomStd. If that does not work, A: ZoomFit resizes y to fit the graph.
reset the window size with  WINDOW.
Inside  WINDOW: F2 Zoom can also be accessed from inside 
WINDOW.
xmin & xmax set the values of x on the left
and right sides of the window, similarly Friendly Windows
for ymin & ymax. A friendly window is a window where the x
xscl (“x scale”) sets the distance between coordinates of the pixel elements are round
tick marks on the x-axis, similarly for numbers, e.g. 1, 2, 3, 4, ... or 0.1, 0.2, 0.3,
yscl. 0.4, ... This is very helpful when using the
Trace feature.
xres sets pixel resolution; 1 = highest
resolution; 10 = lowest resolution; 2 is In  GRAPH, Zoom 4: ZoomDec (decimal)
default. The lower the resolution, the sets the pixel size to 0.1 and the window
faster a graph is drawn. See the dimensions to -7.9 < x < 7.9, -3.8 < y < 3.8.
discussion of friendly windows below. Zoom 8: ZoomInt (integer) sets the pixel
size to 1.0 and allows you to use the arrows
Zoom to move to the center to the part of the graph
To view the equation’s graph, key  GRAPH. you wish to investigate. (ZoomDec does not
Inside  GRAPH, key F2 Zoom to resize the allow this option.) If you choose to center
window (i.e. to change the maximum and / or the graph at the origin, the window size is -
the minimum value of x and / or y that is 79 < x < 79, -38 < y < 38.
displayed). You might want to zoom out so But xres determines how many pixels are
that you can see all of the main features of actually traceable. (Skipping pixels speeds
the graph - intercepts, asymptotes, min/max up the graphing process.) What you actually
and behavior as x ± You might want to trace is 1, 2, 3, 4, ... or 0.1, 0.2, 0.3, 0.4, ...
zoom in so as to precisely determine an x- (for ZoomInt & ZoomDec and respectively)
intercept or to understand a puzzling times xres. If you really want 1, 2, 3, 4, ... or
behavior of the graph. 0.1, 0.2, 0.3, 0.4, ... (normally you do want
Inside  GRAPH F2 Zoom that), set xres to 1. The default is 2.

1: ZoomBox zooms in on a box you draw. The scale set by ZoomInt is often the wrong
size, i.e. you get steps of 1.0, but need steps
2: ZoomIn & 3: ZoomOut zoom in & out by of 0.0001 or maybe 1000, etc. If you need
the amount you set in C: SetFactors. The steps of 0.1, then use ZoomDec. But if the
default is 4 region you want to trace is off the screen,
4: ZoomDec & 8: ZoomInt set friendly you are out of luck with ZoomDec. It is
windows. See the discussion of friendly possible to develop a formula to calculate
windows below. the window size to set to solve this problem,
but that is tedious. Here is a trick. In
GRAPH set F2 Zoom C: Set factors, set
 Tips for Math with the TI-89 Calculator, page 3
xFact and yFact to 10 (or if you prefer a parametric graph is made on the x-y axes by
smaller zoom step size to 10 - the default defining x = f(t) & separately y = f(t). Thus in
zoom step size is 4). Select F2 Zoom 8: parametric mode, you must type in a pair of
ZoomInt. Set the center where you need it. equations. E.g. in Y=
Now Select F2 Zoom 2: ZoomIn twice (or xt1 = sin 2t
once if you set the zoom factor to 10). Now
you have step sizes of 0.1. If you need step yt1 = sin 3t
sizes of 0.01, repeat, etc. To view the graph of the above set x & y
min/max = ±1 & use radian mode. You will
Trace get a pretty Lissajous figure.
Inside  GRAPH, F3 Trace puts a cursor on the You must use t (not x, y or z) as your
graph & displays the coordinates of the cursor. independent variable.
The cursor can be moved along the curve with
◄ or ► or by typing an x value and ENTER.  Window now has (in addition to xmax, yscl,
This can be used for finding intercepts or other etc.) tmin, tmax & tstep, which you may
solutions to the equation, for reading out data need to set.
points in a scatter plot or histogram heights.
Polar Graphs
Other Tips for Graphing To graph in polar coordinates (p. 121-26) key
To change the center of the graph, move the MODE Graph = 3: POLAR.
cursor to the desired center and key ENTER. The Y= screen will now read r1=, etc.
To format a graph (rectangular vs. polar, grid You must use  (not t, x, y or z) as your
on/off, label axes on/off, etc.) key  or in independent variable.  is  (above the ^
Y=, Window or Graph, key F1 9: Format. key).
To cancel a graph while it is being plotted, key Use ZoomSqr to set the correct proportions or
ON. do it by hand by setting xmin & xmax to
The normal graphing mode is MODE Graph = twice ymin & ymax.
1: FUNCTION. Use this e.g. to graph y1(x) =
 Window now has (in addition to xmax, yscl,
x².
etc.) min, max & step, which you may
You can store a window setting with F2 Zoom need to set.
B: Memory 2: ZoomSto. You can recall your
If some functions are selected, they might
stored setting with B: Memory 3: ZoomRcl.
graph along with your polar graph. To turn
In  GRAPH to find the Minima or Maxima them all off, in Y= key F5 ALL 1: All Off,
of a function use F5 Math 3: Minimum or 4: 3: Functions Off or 5: Data Plots Off as
Maximum. You will be prompted to choose an needed.
x value on each side of the zero.
 | Coordinates  Polar will cause F3 Trace
You can depict an inequality (e.g. y > 2x + 3) to display the coordinates r & .  |
on the x-y plane by graphing the inequality as Coordinates  Rectangular will cause F3
an equality solved for y (e.g. y1(x) = 2x + 3) Trace to display the coordinates x, y & 
and then shading above or below the graphed
line, depending on whether the inequality was To find the x and y Intercepts
actually > or <. (You want to shade above in
To find x and y intercepts, graph the equation
our example.) To set the shading in Y=, key
and zoom in or out with F2 Zoom (or better
F6 Style 7: Above or 8: Below as needed.
by using  Window reset x & y min & max
There are 4 shading patterns which are
until the x and y intercepts can be seen on
automatically cycled through. So 4 different
the screen). To use the following method the
inequalities can be displayed. See p. 100.
intercepts must appear on the screen.
Parametric Graphs To find the y-intercept, key F3 Trace, then
To make a parametric graph (p. 115-20) key 0, then ENTER. This will move the cursor to
MODE Graph = 2: PARAMETRIC. A x = 0, i.e. the y-intercept. The y-intercept
Tips for Math with the TI-89 Calculator, page 4
will be displayed in the lower right of the there and connects the dots. Thus where an
screen. asymptote should appear, a slightly crooked
fake asymptote might be drawn.
To find the x-intercept, key F5 Math, then
2: Zero. You will be asked for the “Lower If you want to get rid of fake asymptotes, in
Bound?”. Either place the cursor to the left Y= set the style to dot. Or set the scale so that a
of the x-intercept with the ◄ key and key pixel element falls on the asymptote. E.g. for y
ENTER or if you know roughly the value of = 1/(x-1) there is a vertical asymptote at x = 1.
the x-intercept, key in an x value less than F2 Zoom 4: ZoomDec and  WINDOW xres =
the value of the x-intercept. Then you will 1 will put a pixel element at x = 1 and thus no
be asked for the “Upper Bound?” Either fake asymptote will appear.
place the cursor to the right of the x- If you want to draw an asymptote. If you
intercept with the ► key and key ENTER or know where an asymptote is, you could key its
key in an x value greater than the value of equation in - perhaps in a different style and add
the x-intercept. it to your graph. Horizontal asymptotes can be
easily keyed in. E.g. a horizontal asymptote at y
Tables = 2 would be graphed as y2(x) = 2.
To make a table (p. 210-218), create a function Vertical asymptotes can only be approximated,
in Y=. TABLE produces a table of the e.g. for a vertical asymptote at x = 2, use y2(x)
functions selected in Y =. Depending on the = 10^100(x-2), which is a line which goes
MODE setting, this could be functions y1, y2, through the point (2, 0) and has such a steep
etc, or parametric functions y1t, x1t, etc, or slope that it will appear perfectly vertical.
polar angle , etc. Below I assume you are in Alternatively type y2(x) = when(x<2,-10^100,
function mode. 10^100). This draws a horizontal line at -10100
for x < 2 and a horizontal line at 10100 for x  2.
There is a choice of ways to choose which x Both are way off screen. What you see is the
values to display. In TABLE SETUP, which vertical line connecting these two lines.
can be accessed with  TblSet or inside Alternatively type LineVert 2 in the home
TABLE with F2 Setup AUTO table screen, not in Y=. You cannot trace or find
automatically generates a series of values for x intersections, etc. of a line drawn with LineVert.
or you can choose them yourself with ASK.
For example key y1(x) = sin x. Deselect any To delete functions, lists, tables,
previously selected functions. Set the table etc.
parameters with tblStart = -90, tbl = 15
(assuming you are in degree mode & want a To delete tables, etc. key 2nd VAR-LINK. p.
table of sin x for every 15° starting at -90°). 333-41.
Graph <-> Table = OFF (unless you want to All of your (user defined) variables: tables,
use xmin & xres to set tblStart & tbl, which functions, lists, text entries are listed.
is not recommended). Independent = AUTO.
ENTER Highlight the table, etc. you want to delete. Key
 Enter.
If you want to choose your own values of x, use
Ask. E.g. in  TABLE key F2 Setup, If you want to delete several variables, select all
Independent  Ask, ENTER. Then key in to be deleted with key F4 , then key  Enter
your x value, ENTER, , then key in another x as above.
value, etc. If you want to delete all but a few variables, key
To change the cell width in TABLE key | or F5 All 1: Select All. Then use F4  to deselect
F1 9: Format those tables you do not want to delete.
By default all user defined variables are placed
Asymptotes in the MAIN folder unless you specified
The TI-89 does not draw asymptotes, but otherwise. You cannot delete the MAIN
because of the way it draws a curve, fake folder.
asymptotes sometimes appear. It evaluates y
at the center (in x) of each pixel, draws a dot
Tips for Math with the TI-89 Calculator, page 5

Converting from Rectangular to displayed depends on what modes you are in.
For example
Polar Coordinates or Vice Versa
In approx mode 2ei/4ENTERgives +
To convert from Rectangular to Polar
i.
Coordinates key [x,y] 2nd MATH 4: Matrix L:
Vector ops 4:  Polar In exact mode 2ei/4ENTER gives + 2i.
For example [1, 3)]:  Polar gives [2  /3] The Solve & Zeros commands
To convert from Polar to Rectangular To solve an equation key F2 Algebra 1:
Coordinates key [r,] 2nd MATH 4: Matrix solve( in the entry line. E.g. solve(2x + 3 =
L: Vector ops 5:  Rect 7,x) gives x = 2. This single line will do about
half of your homework or test problems! If I
For example [2,  /3]  Rect gives [1 3]
want you to do the problem yourself by hand,
The “ ” is 2nd EE. The “,” does not display in I will require you to show your work or I will
the answer, but it is there. See p. 456, 464. prohibit calculators on part of the test.
Converting Complex Numbers Some equations cannot be solved analytically
(i.e. with algebra), e.g. x = cos x. With a
from Rectangular Form to Polar grapher there are several ways to solve it.
Form or to Trigonometric Form, One way is graphically by plotting y1 = x, y2 =
etc. cos x and in  GRAPH find the intersection
To convert a complex number in standard of two functions using F5 Math 5:
(i.e. rectangular, i.e. a + bi) form to polar intersection. You will be prompted to choose
(i.e.. r ei) form, set the mode to complex two functions and an x value on each side of
polar with MODE Complex Format 3: the intersection.
POLAR, key in the complex number in Or in the entry line use solve(x = cos(x), x),
standard form & press ENTER. The way the giving “x = .739085133”.
result is displayed depends on what modes
you are in. You may need to reset the Or set the equation equal to zero (e.g. x - cosx =
Exact/Approx mode or the Angle mode (i.e. 0) and in the entry line use zeros(x - cos(x),
Degree vs. Radian mode). For example: x), giving “x = .739085133”. Zeros is F2 4:
zeros(.
In exact, radian mode + 2i ENTER gives 2
Or graph y1 = x - cosx and find the zeros of the
ei/4.
graph with F5 Math 2: Zeros. You will be
In approx, radian mode + 2i ENTER gives prompted to choose an x value on each side of
2 e0.785 i. the zero. For a more detailed explanation of
how to use the zeros feature, see below.
In exact, radian mode 5 + i ENTER gives ei
arctan5/5
3 2. The above two Zeros functions are different.
The one used in the entry line does not ask for
In approx, polar, degree mode + 2i ENTER a range to search and might not find all zeros.
gives 2.44924.095.
Of all of the above methods, I recommend
Etc. using intersection in the GRAPH screen,
Apparently the TI-89 cannot be commanded to because by looking at the graphs you can
display in trigonometric form (e.g. 2 cos 45° + understand what is going on, because if there
2i sin 45°). is more than one intersection you can choose
the one you want and because setting the
To convert a complex number in polar (i.e. equation equal to zero or equal to y involves
rei) form to a complex number in standard extra algebra and is therefore not as intuitive.
(i.e. rectangular, i.e. a + bi) form, set the
mode to complex rectangular with MODE Either Solve() or Zeros() can be used for most
Complex Format 2: RECTANGULAR, key in problems. For example Solve(x² - 4 = 0, x) or
the complex number in standard form and Zeros(x² - 4, x) will both give 2 & -2.
press ENTER. Again the way the result is
Tips for Math with the TI-89 Calculator, page 6
There are three Solve() & two Zeros() 2i) / n = (/2)/3 = /6). Set step to 2/n =
commands: Solve() p. 478-481, Zeros() p. 2/3. Set the window size appropriately, i.e. a
494, cSolve() p. 403-405, cZeros() p. 407 and little bigger than ³2 e.g. ±2. Key GRAPH.
nSolve() p. 450. Key F2 Zoom 5: ZoomSqr (to make the plot
If you want complex solutions, you must use “circular”). Key F3 Trace. The first root,
cSolve() or cZeros(). Solve() or Zeros() or 1.091 + .630 i will be displayed (as
nSolve() will give only the real solutions. 1.091, .630). Key ► to display the second root
-1.091 + .630 i. Key ► again to display the
Solve() is 2nd Math 9: Algebra 1: Solve(). third (& last) root 0 - 1.260 i.
Zeros() is 2nd Math 9: Algebra 4: Zeros(). There are at least five possible ways (!) that
nd
nSolve() is 2 Math 9: Algebra 8: nSolve(). the results of cSolve() or cZeros() can be
displayed. E.g. for x² = 4i, i.e. finding the
cSolve() is 2nd Math 9: Algebra A: Complex 1:
square roots of 2i, the answer could be
cSolve().
displayed in exact polar radian form ei, e.g.
cZeros() is 2nd Math 9: Algebra A: Complex 3: 2ei/4 and 2ei/4 or approx polar radian form
cZeros(). ei, e.g. 2 e 0.785 i and 2e -2.366 i or polar degree
nSolve looks for only one approximate real form (r  ), e.g. (2  45) and 2  -135) or
solution, but is faster than Solve() or Zeros(). exact standard form a + b i, e.g.  + 2 i
and -2 - 2 i, or approximate standard
Solve is very powerful.
form a + bi, e.g. 1.414 + 1.414i & -1.414 -
 It can solve symbolic equations, e.g. 1.414i.
Solve(ax^2 + b*x + c = 0, x) will give the Apparently the TI-89 cannot be commanded to
quadratic formula. display the results in trigonometric form (e.g.
 It can find the intersection of two equations, 2 cos 45° + 2i sin 45°).
e.g. Solve(x^2 + y^2 = 4 and y = x^2, {x, y}) Which form is displayed is controlled by 3 (!)
will give (±1.250, 1.562) i.e. the intersections mode settings. Mode: Complex Format 3:
of the circle and the parabola. (“and” is 2nd Polar will always display the result in polar
Math 8: Test 8: and.) form.
 Solve(x4- 1 = 0, x) will give x = 1 or x = -1, Mode: Complex Format 1: Real or 2: Rect will
but cSolve(x4- 1 = 0, x) will give x = i or x = -i always display the result in standard form.
or x = 1 or x = -1.
Mode: Exact/Approx 2: Exact will always
 To speed up the process, you can also display the result in Exact form, i.e. ei/4 or
include a guess e.g. Solve(x^2 = 2, x = 1.4) 2/2 + 2/2 i, depending on the Complex
will give 1.414, although the equation must be Format mode.
more complicated before the speed-up is
noticeable. Mode: Exact/Approx 3: Approximate will
always display the result in Approximate
Solving with Complex Numbers form, i.e. e 0.785 i or .707 + .707 i, depending on
cSolve() or cZeros() can be used to find the the Complex Format mode.
roots of complex numbers. For example ³i In Angle 1: Radian mode the polar radian form,
= cSolve(x³ = i, x) = cZeros(x³ - i, x) = {-i, - r ei, is displayed, i.e. ei/4 . In Angle 2: Degree
3/2 + ½ i, 3/2 + ½ i} mode the polar degree form, (r  ) i.e. (1 
DeMoivre’s Theorem can be used to display 45), is displayed .
and trace (i.e. read out) the roots of Complex numbers are explained on p. 533-535.
complex numbers. For example DeMoivre’s
Theorem says that the cube roots of 2i = 2 ei/2 How to Find the Intersections of
are ³2 ei(/2+2k)/3, with k = 0, 1, 2. To graph Two Graphs
these values set the Graph mode to Polar. In Enter the two equations in Y=. Select them and
Y= key 2^(1/3). In WINDOW set min to the deselect all other equations. Display them with
angle of the first root (i.e. (the argument of  GRAPH. Key F5 Math 5: intersection. You
will be prompted to choose the two functions.
Tips for Math with the TI-89 Calculator, page 7
If (as instructed above) you have only selected E.g. to display the first 6 terms of the arithmetic
the two equations whose intersections you sequence an = 3n + 2, key seq(3x+2, x, 1, 6)
wish to find, press ENTER ENTER. If not, ENTER which gives {5 8 11 14 17 20}. Note I
use the up or down arrows to move the cursor used “x” rather than “n” as my variable to
to the graphs whose intersections you wish to make the keying easier. seq(3x+2, x, 3, 5, ½)
find and press ENTER for each graph. Then ENTER gives {11 25/2 14 31/2 17}.
you will be prompted to choose an x value on To find the sum of a sequence you can use
each side of the intersection. In response to
sum(seq( (See p. 483.) or ( (See p. 507.).
“Lower Bound?”, you can either key in an x
value less than (left of) the x-coordinate of the The sum &  features are different. Sum just
intersection (If you know the size of your takes a list to sum, for example sum({1, 2, 3})
screen, you can figure it out.) or you can use gives 6. ( takes parameters (expression,
the left or right arrows to move the cursor to variable name, begin, end). E.g. to find the
the left of the intersection. Press ENTER. sum of the first 4 terms of the geometric
“Lower” is a bit confusing. Remember it sequence an = 52n, key sum(seq(5*2x, x, 1, 4)
means “left”, not “below”. In response to ENTER or key (5*2x, x, 1, 4) ENTER, either
“Upper Bound?”, similarly choose an x-value of which give 150. Since ( is the same as
to the right of the intersection. To move the sum(seq(, it’s easier to just use (.
cursor faster, use 2nd left or right arrow.
But ( does not accept a step size, i.e. the step
Factoring and Expanding size is set to 1. If another step size is needed,
To factor an expression key F2 Algebra 2: e.g. for Riemann sums, then you must use
factor( in the entry line. E.g. factor(x^2 - 5x + sum(seq(. For example for f(x) = 2x to find
6, x) gives (x-3) (x-2). f(2) + f(2.2) + f(2.4 + ... +f(4), use
E.g. factor(x³ + 3x²y + 3xy² + y³) ENTER sum(seq(2x,x,2,4,.2)), which gives 96.700.
gives (x + y)³. Or you can get clever with (. To find the sum
E.g. cfactor(x4-1) gives (x-1) (x+1) (x-i) (x+i) of the n terms of f(x) from x = a to b (i.e. with
step size =x = (b-a)/n use
To expand an expression key F2 Algebra 3: (f(a+xx),x,0,n). For example for f(x) = 2x
expand( in the entry line. E.g. expand(x + to find the sum of the 10 terms f(2) + f(2.2) +
y)^3) ENTER gives x³ + 3x²y + 3xy² + y³.
f(2.4) + ... +f(4), x = (4-2)/10 = .2, use (22+.2
x
Permutations and Combinations , x, 0, n), which, of course, also gives 96.700.

Combinations are the number of ways of The result can contain symbols, e.g. (x,x,1,n)
selecting r objects from n objects where order ENTER gives n(n+1)/2.
does not count = nCr = n! / r! (n-r)! is also To find the partial sums of a sequence use
written as . cumSum(. (See p. 405.) E.g. key 2nd MATH 3:
List 7: cumSum(seq(1/x², x, 1, 4)) ENTER
gives {1 5/4 49/36 205/144}
Permutations are the number of ways of
selecting r objects from n objects where order A sequence graph is will display a sequence.
does count = nPr = n! / (n-r)! Set MODE Graph SEQUENCE. Key  Y=.
Cr is 2nd MATH 7: Probability 3: nCr( Enter the formula for the sequence in u1. If an
n
initial value is needed, e.g. if a recursive
P is 2nd MATH 7: Probability 2: nPr(
n r formula for the sequence is entered in u1,
enter the initial value in ui1, otherwise leave it
Sequences and Series blank. E.g. key in u1 = 3n+2. You must use
A sequence is an ordered list of numbers. A “n” as your variable in sequence mode. Key 
series is the summation of the terms of a GRAPH F2 Zoom A: ZoomFit. The axes are n
sequence. They can be calculated using 2nd and u1. You can use F3 Trace to read out the
MATH 3: List 1: seq(. (See p. 470, 127-139.) values of u1 (they are called yc).
Seq takes the parameters (expression, variable
name, begin, end [, step]). Step is optional, the How to graph a Conic Equation
default is 1.
Tips for Math with the TI-89 Calculator, page 8
Since a conic equation is not solved for y, it Solving a system of two equations, i.e. finding
cannot be directly entered into Y=. But by the intersections(s) of their graphs.
treating x as a constant, a conic can be Solve the 2 equations for y. Graph them
rearranged into a quadratic equation in y. Then
using the  Y= and  GRAPH. If the
the quadratic formula can be used to solve for y.
intersection(s) do not appear on the screen,
The general form of the conic, Ax2 + Bxy + Cy2
zoom out (or better by using  Window
+ Dx + Ey + F = 0, can be rearranged to Cy2 +
reset x or y min or max until the
(Bx + E)y+(Ax2 + Dx + F) = 0. Note that this
intersection(s) can be seen on the screen).
now has the form of a quadratic in y. Applying
the quadratic formula gives Key F5 Math 5: intersection. In case you
have other functions displayed on the screen
. besides the ones you want to solve, you will
There is one problem with keying in this be prompted with “1st Curve?” and then “2nd
formula, there is no “±” button. The solution is Curve?” to choose the two functions whose
easy, the list {1, -1} is effectively a “±” sign. If intersections you want to find. If the cursor
you have one conic to graph, key the above is on the correct 1st & 2nd curves, just press
formula in for y1= with the letters (A, B, C, ENTER to answer each question, otherwise
etc.) replaced with their values. use the ▲ or ▼ key to move to the correct
curve(s). Or better still, use Y= and then F4
For example for x² - 4xy + 4y² + 10x - 30 = 0 to uncheck the unneeded curves.
(i.e. A = 1, B = -4, C = 4, D = 10, E = 0, F = -
30), key in y1 = (4x+{1,-1}((-4x)^2- Then you will be prompted “Lower Bound?”
Use the ◄ key to move the cursor to the left
4*4(x^2+10x-30)))(2*4).
of the intersection and key ENTER or if you
If you want to graph several conics, it is faster know roughly the x-coordinate of the
to create a function called conic(x) which intersection, key in an x value less than the
contains the above formula, i.e. in the home x-coordinate of the intersection. You will be
screen key prompted “Upper Bound?” Use the ► key to
(-(b*x+e)+{1,-1}*((b*x+e)^2- move the cursor to the right of the
4*c*(a*x^2+d*x+f))) (2*c)  conic(x), where intersection and key ENTER or if you know
roughly the x-coordinate of the intersection,
 is the STO button. Note that you must
key in an x value more than x-coordinate of
explicitly type b*x, not bx, otherwise bx will be
the intersection.
interpreted as a variable name, not b times x.
Then for each conic just key in the values of A, Then the (x, y) coordinate of the intersection
B, C, etc. will be displayed at the bottom of the screen.
For the above example, key 1  A, -4  B, 4 Define
 C, 10  D, 0  E and -30  F. Now set
y1(x) = conic(x), and graph it. F4 Other 1: Define can be used to define a
function. E.g. Define xxt(x) = 3x^2. Then
How To Simplify Rational xxt(5) ENTER gives 75. Notice that I used
Functions letters available on the keyboard to make
typing the name faster. You would have gotten
To do long division of one polynomial by an error message if you had tried to define
another, use expand(. For example xt1(x), because xt1(x) is a system variable.
expand((x^2+4x+6)/(x+2), x) gives There is another way to define a function.
2/(x+2)+x+2. Expand is F2 Algebra 3: expand(. 1cos(x) STO sec(x) defines a new
MODE Split Screen 3 LEFT-RIGHT or 2: function, sec(x) which can, for example, be
TOP-BOTTOM is useful with graphs. Make graphed by entering y1 = sec(x) or evaluated
one screen the graph and use the other screen by typing sec() which gives -1. An erroneous
to adjust the window and to set the style (dot, set of keystrokes can result in defining x & y.
thick, etc.) of the graph. This is surprisingly easy to do by accident.
E.g. if x has been defined as 2, expand((x+2)²)
Solving a system of two equations will give 16. If something like this happens,
F6 Clean Up 1: Clear a-z ENTER ENTER
Tips for Math with the TI-89 Calculator, page 9
will clear (i.e. delete) the definitions of any 1-
character variables, i.e. “x”, but not “xx”. see For example can be written as,
p. 85, 192 y10(x) = when (x > 1, x, x^2).
If you want to use your newly-defined function,
easier than keying in the letters is copying a
To write , you must use
variable from VAR-LINK. Key 2nd VAR-
LINK. Highlight the variable (e.g. the sec(x)
function created above) ENTER, sec(x) now nested when(s, For example:
appears in the entry line. y1(x) = when (x  1, when(0 < x, 3, -x), x) or
Piecewise-Defined Graph y1(x) = when (x  0, -x, when(x 1, 3, x))
To display a piecewise-defined graph, e.g. Notice the difference in the inequalities used
above and that the -x + x term is no longer
, needed.

inside Y= key For example 4 - x², x < 1

y1 = when(x > 1, x, else x^2) f(x) = 1.5x + 1.5, 1  x  3
when( is in CATALOG.
x + 3, x > 3
If the graph has more than 2 pieces, e.g.
is entered as y1 = when(x<=3, when(x < 1,
4 - x^2, 1.5x + 1.5), x + 3)
, Once you have defined a function in Y =, you
can use it in the entry line. E.g. with the above
you can use nested when functions or a user- definition, y1(-25) evaluates as 25
defined function, see p. 194-95. Unfortunately To display f(x)={x, x  1, key y1 = when(x  1,
the logic of nested when functions is hard to x, undef). undef means undefined, so it won't
follow, especially since you have to read it draw anything if x > 1. Just type “undef “ in,
inside the small entry line. it’s not in the catalog!
It is easier is to enter a piecewise defined
equation as separate equations, then selecting
The Inverse
each with F4 . E.g.: To find analytically the inverse of f(x), written
f-1(x), key solve(y = f(x), x). Then exchange x
y1(x) = -x | x  0
and y by hand. Solve( is F2 Algebra 1: solve(.
y2(x) = 3 - x + x | 0 < x and x  1 For example solve(y = 2x + 1, x) gives x = (y-
1)/2. So f -1(x) = (x-1)/2. Solve(y = ln(x), x)
y3(x) = x | 1< x
gives x = ey. So f -1(x) = ex.
The unexpected “- x + x” term was used to get
There are several ways to graph the inverse.
an x into the expression. Without an x in the
expression “y2(x) = 3 |0 < x and x  1” would 1) You can first find f-1(x) analytically as above
have drawn a 3 for all x, not just for 0 < x  1. and then just key it in in Y =. For example y =
x² Solve it for y, i.e. y = ±x. This can be
The Boolean operator and is 2nd MATH 8: Test
graphed as y ={1,-1}x. The list {1,-1} is
8: and or CATALOG and.
effectively a ± symbol.
Unfortunately if you want to then graph g(x - 3)
2) Or use parametric mode:
or if you need a table of g(x) or if you want to
find values e.g. g(-3), then you need to express xt1 = t
the function in one entry. If so the g(x) above yt1 = f(t)
can then be defined using when(test,
expression when test true, expression when will produce a graph of f(x).
test false). xt2 = yt1(t)
yt2 = t
Tips for Math with the TI-89 Calculator, page 10
will produce a graph of f-1(x). considered to be defined only for x  0, but in
For example, xt1 = t, yt1 = t^2, will produce a Complex Format “real” mode the TI-89 shows
graph of y = x², xt2 = yt1(t), yt2 = t, will a graph of y = x2/3 (but not of y = x0.7) for all x.
produce a graph of x = y², its inverse. If you want the “normal” graph of y = x2/3, i.e.
for x  0 only, set the TI-89 to Complex
You may need to adjust the values of tMin and Format "rectangular" or "polar" mode.
tMax to allow for the range of y values you
want. Set in Complex Format "rectangular" or
"polar" mode the TI-89 finds the "first"
3) Or type on the command line DrawInv x^2, root. For example, if x = -8, it returns x2/3
which draws the inverse of y = x². = -2 + 23 i, which it cannot graph. Set in
All of the above is for the inverse relationship Complex Format "real" mode the TI-89
i.e. where you do not require that the inverse finds the first real root. For example if x =
be a function. If you want the inverse -8 it returns 4, which it can graph. Another
function and the inverse relationship is not example is y = . Since is
already a function, it does not work.
To obtain the graph of the inverse function imaginary for x < 4, this function is
use Method 1. If the function did not have an considered to be undefined only for x  4,
inverse function because it contained a y² term but the TI-89 does the calculation using
(e.g. y = x² or x² + y² = 1), proceed as in 1 complex numbers. Since is
above, but remove the {1, -1}. imaginary for x < 2, the quotient is real

If f(x) contains a transcendental function, (i.e. and the TI-89 graphs y = for x  2
one which cannot be inverted algebraically, as well. I do not know how to make the
e.g. tan x), use the inverse transcendental, e.g.
TI-89 refuse to graph this for x  2.
tan-1 x. For example solve(y = 2 tan x + 3,x)
gives x = {tan-1[(y-3)/2]}/ + @nj. (The j Linear Interpolation
stands for an integer. @nj is the name of an
If you have a table of data, e.g.:
integer constant. If the solution has two
x y = f(x)
constants, they will be named @n1 and @n2.),
4 24.7
i.e., in more standard notation, x = {tan-1[(y-
5 25.2
3)/2]}/ + k, k = an integer. 6 25.8
To get the inverse function just drop the + @nj and wish to find a y-value corresponding to an
part. So the inverse function is f-1(x) = tan-1[(x- x-value between those x-values in the table,
3)/2]/ In this example DrawInv 2 tan x + 3 assume the (unknown) function y = f(x) is
or graphing y1 = {tan-1[(x-3)/2]}/ + @nj, locally linear and use linear interpolation. Use
gives a series of curves, one on top of the the following equation, where x1 and x2 bracket
other. On the other hand graphing y1 = {tan-1
[(x-3)/2]}/ gives just the one curve of the x.: y = y1 +
function, called the principal branch.
E.g. find f(5.8) = 25.2 + (25.8-25.2)(5.8-5)/(6-5)
Graphing a Complex Function = 25.68
A complex function, z = x(t) + y(t)i, can be
Programming
graphed on the complex axis in parametric To type in a program, key APPS 7: Program
mode as xt1 = x(t) and yt1 = y(t). Or if z is not Editor 7: Program Editor 3: New Type:
explicitly written with real and imaginary Program. Give it a name, e.g. Variable:
parts, i.e. just as z = f(t), use xt1 = real(f(t)) example.
and yt1 = imag(f(t)).
The lines
Graphing a Real Function :example()
:Prgm
The domains of some functions are restricted, :
because they produce complex results for :EndPrgm
some x. For example y = x0.7 and y = x2/3 are
Tips for Math with the TI-89 Calculator, page 11
then appear automatically. Note the first line is Decimal means base 10, binary means base 2 &
the name you assigned to the program. hexadecimal means base 16.
How do you tell the calculator you are entering
Type in your program between Prgm and
a binary or hexadecimal number? You type 0b
EndPrgm. Key ENTER after you finish each
before a binary number & 0h before a
line. When you are done typing the program,
hexadecimal number. For example to enter 7 in
key HOME.
binary, key 0b111. To enter 15 in hexadecimal,
To save on typing key F3 I/O to insert “Disp”, key 0hF.
“Input”, “Output”, etc. into your program. How do you set the mode of the answer of a
To run the program, type its name in the entry calculation? To display your answer in binary,
line on the home screen, e.g. example(). Note key MODE, Base, 3: BIN. To display your
that the () were not used above when naming answer in hexadecimal key MODE, Base, 2:
the program. HEX. To display your answer in decimal (i.e.
the normal way) key MODE, Base, 1: DEC.
Lbl creates a label which the program can later You can enter the number in any mode, the
jump to with a Goto statement. answer will be displayed in the chosen mode.
Input [prompt], variable prints the prompt on For example in decimal mode 0b111 + 0b1
the screen. After you type in a value for the gives 8. In binary mode 0b111 + 0b1 gives
variable and key ENTER, that value is 0b1000. In hexadecimal mode 0b111 + 0b1
assigned to the variable. gives 0h8. You must always use the prefixes.
For example in hex mode 11 + 3 gives 0hE, i.e.
Disp displays the current contents of the 11 is interpreted as decimal.
Program I/O screen. How can you convert a number from one base
To leave the Program I/O screen, i.e. stop a to another? There are two ways. Firstly you can
running program, key F5 PrgmIO or key ON set the mode to the desired base and key in the
ENTER HOME. number with the correct prefix. For example in
decimal mode 0b111 ENTER gives 7. In hex
To Solve a System of Equations mode 15 ENTER gives 0hF.
To solve a system of three equations and three Alternatively you can use the convert base
unknowns key solve(eqn1 and eqn2 and eqn3, commands in 2nd MATH D: Base, which are
{var1,var2,var3}). 1:Hex, 2:Bin & 3:Dec. For example,
0b11010Hex gives 0h1A and 0b11010 Dec
E.g. to solve x + 4y - z = 6, 3x + 2y + 3z = 16
gives 26. These commands only work in
and 2x - y + z = 3, key solve(x+4y-z=6 and
EXACT or AUTO, not APPROX mode.
3x+2y+3z=16 and 2x-y+z=3,{x,y,z}), which
gives x=1 and y=2 and z=3.
On Line Help
Step functions TI provides dozens of forums on its TI
[x] = the greatest integer less than or equal to x Calculators, for example there are forums
= int(. This function is also called floor(. entitled “TI-89”, “Precalculus”; “Calculus”, etc.
For example, int(3.5) returns 3 and int(-3.5) Look at links below
returns -4. http://education.ti.com/index.html
If you want the smallest integer greater than or or
equal to x, use ceiling(. http://education.ti.com/student/
For example, ceiling(1.5) returns 2. ceiling(-1.5) TIStudentCenterHome.html
returns -1. or
If you want the integer closest to zero, i.e. http://www-s.ti.com/cgi-bin/discuss/
simply removing any fractional part, use sdbmessage.cgi?
iPart(. databasetoopen=calculators&topicarea=TI-
For example, iPart(1.5) returns 1 and iPart(-1.5) 89/92+Plus&do_2=1
returns -1.
These can be found in the MATH 1: Number.
Binary, Hexadecimal & Decimal
Error Messages
Tips for Math with the TI-89 Calculator, page 12

Error: Dimension quotient if x is kept finite. The default

value of x is 0.001.
Error: Dimension means that the variable you
are using, e.g. x, is not the correct dimension for E.g. nDeriv(x4,x,.1) calculates [f(x+.1)-
the function you are using, e.g. Factor(x^2 - 9). f(x-.1)]/.2, which gives 4 x (x² + .01),
This is probably because by accident you stored whereas d(x4,x) gives 4x³, i.e. the exact
something in x. Either use F4 Other 4:DelVar x derivative.
or F6 Clean Up 1: Clear a-z.
Integration
2nd (expression, variable[, lower][, upper]) (on
Calculus the 7 key) [or equivalently F3 Calc 2:
( integrate] does integration.
E.g. x²,x) ENTER gives x³/3
The two main operations of calculus are E.g. x cosx,x) ENTER gives cosx + x sinx
differentiation and integration, of course.
Both of these can be done easily either E.g. 1/x,x,1,2) ENTER gives ln2
numerically or symbolically. They can be lower is added as a constant of integration if
done on the home screen with the keyboard upper is omitted. E.g. x²,x,c) ENTER gives
or on the graph screen. In addition to x³/3 + c
numerical differentiation and numerical
integration, the AP Calculus Course F3 Calc has other useful operations.
Description lists two other calculator
operations that are required for the AP
Limits
Calculus exam: “plot the graph of a function F3 Calc 3: limit( will find a limit.
within an arbitrary viewing window” and Syntax: limit(expression,var,point[,direction]).
“find the zeros of functions (solve equations The optional direction parameter indicates
numerically).” Setting the window size is from which side a one-sided limit is taken. If
covered in my Tips for Beginners for the TI- direction is a positive number, the limit is
89 Calculator from the right, if direction is a negative
number, the limit is from the left.
On The Home Screen
E.g. limit(3x/(1-2x), x, ) ENTER gives -2/3.
Differentiation e.g. limit(1/x, x, 0) = undef
2nd d (expression, variable[, order]) (on the 8 E.g. limit(1/x, x, 0, 1) =  (from the right side)
key) does differentiation.
E.g. limit(1/x,x,0,-1) = - (from the left side)
E.g. d(x³,x) ENTER gives 3x².
Appendix A has further examples.
E.g. d(f(x)g(x),x) ENTER gives the quotient
rule (although in a nonstandard form). Finding Epsilon in the Limit
To find f’(a) [where a is a number] use “|” E.g. Definition
d((x-1)(x+1)),x) |x = 3 ENTER gives 1/8.
L= iff for any , there is a  such that
Order is optional, the default is 1. For a 2nd
derivative use 2. E.g. d(x³,x,2) ENTER gives if x is within  of c, f(x) is within  of L. So
6x. If order is negative, the result is an L ±  = c ± .
antiderivative. E.g. d(x³,x,-2) ENTER gives To find , use solve(. Key in solve(L +/-  = f(c
x5/20, note the constant is dropped. + ), ). You have to key in the + & then the
Numerical Differentiation - separately. You know c, L & , so key
them in as numbers. Since  is the unknown,
d( [the differentiation function above] takes the key it in as “x”. Whichever is the smaller x,
difference quotient with the lim x0. The is . One of the x’s may come out as a
function nDeriv(expression, var [, x] negative number, but ignore the minus sign.
allows the user to explore the difference
Tips for Math with the TI-89 Calculator, page 13
For example: For f(x) = x³ and c = 2, find delta E.g. graphing y = 1/x and in response to the
prompts keying in 1 ENTER and then 2
such that epsilon = 0.1. L = = 8.
ENTER gives .69315.
So key in solve(8+/-.1=(2+x)^3,x). Key in
the + & then the - separately. The - gives x = Partial Fraction Decomposition
.0083683. The + gives x = .0082989. Use expand( for partial fraction decomposition.
Since .0082989 is smaller,  = 0.00829. expand( is F2 Algebra 3:expand(
Note that I rounded .0082989 down, (i.e.
“incorrectly”), because rounding properly E.g. expand(1(x^2-5x+6)) gives
gives a slightly too large , i.e. gives an 
larger than 0.1. Minima, Maxima, Inflection
Minima and Maxima Points, Tangent Lines & Arc
F3 Calc 6: fMin( and 7: fMax( will find a Length
minimum and maximum respectively F5 Math has several other useful operations.
between Lower and Upper Bounds.
3: Minimum and 4: Maximum will find a
E.g. F3 Calc 6: fMin((x-2)² + 3, x) ENTER will minimum and maximum respectively
find the value of x where y = (x-2)² + 3 has a between Lower and Upper Bounds.
minimum i.e. x = 2.
8: Inflection will find an inflection point
Arc Length between Lower and Upper Bounds.
F3 Calc 8: arcLen( will find the arc length. A: Tangent will draw a tangent line at the
entered point and display its equation.
E.g. F3 Calc 8: arcLen( (1-x²,x,-1,1) ENTER
gives 3.14159. B: Arc will find the arc length between “1st
Point” and “2nd Point”.
On The Graph Screen Other options are available in other modes. For
example, in polar mode 6: Derivatives
Plot the graph as usual. I recommend finding
minima and maxima and numerical offers the options of 1: dy/dx or 4: dr/d. In
differentiation and integration, etc. be done parametric mode 6: Derivatives offers the
on the graph screen, because then you can options of 1: dy/dx or 2: dy/dt or 3: dx/dt.
see what’s going on, instead of plugging Convergence of a Sequence
away blindly.
Convergence of a sequence can be displayed.
Differentiation E.g. Key in u1(n) = -.8u1(n-1) + 3.6, ui1 = -
To find the derivative key F5 Math 6: 4 and  GRAPH. The sequence can be seen
Derivatives 1: dy/dx. Then you will be to converge to 2. Or Key  TABLE to see
prompted “dy/dx at?” Either key in the the values. Or simply sum the entire
desired x value and key ENTER or move the sequence.
cursor to the desired x value and key E.g. try (1/x², x, 1, ) ENTER. Surprising?
ENTER.
Taylor Series Approximations
E.g. graphing y = (x-1)(x+1) and in response
Taylor series approximations can be
to the prompt keying in 3 ENTER gives .125
displayed. E.g. sinx = ((-1)^n * x^(2n+1) /
Integration (2n+1)!, n, 0, )). Plot the first 6 terms of
the Taylor series approximation, i.e. y1 =
To find the integral key F5 Math 7: f(x)dx.
((-1)^n * x^(2n+1)/((2n+1)!), n, 0, 5)) and
You will be prompted for the “Lower
also y2 = sin x on the same screen. The
Limit?” Either key in the desired x value and
value of x where the approximation diverges
key ENTER or move the cursor to the
from sin x is very striking. Or to find the
desired x value and key ENTER. You will
numerical value of the differences define
be prompted similarly for the “Upper
y3(x) = y1(x) - y2(x), deselect y1 & y2, key
Limit?”
 TABLE. Use MODE Display digits 9 &
Tips for Math with the TI-89 Calculator, page 14
| to set the cell width to 12 (for the (b-a)/n(f(a+x(b-a)/n),x,0,n-1) STO
maximum number of decimal places) and lreman(a,b,n)
use F2 Setup to select where the table starts RM:
and the step size of the table. (b-a)/n(f(a+(x+½)(b-a)/n),x,0,n-1)
taylor() performs a Taylor series expansion. STO mreman(a,b,n)
The parameters are taylor(expression, RR:
variable, order[, point]), where expression is (b-a)/n(f(a+x(b-a)/n),x,1,n) STO
the expression to be expanded. Point is the rreman(a,b,n)
point about which the series is expanded. Its Trapezoid rule:
default value is zero. The Taylor polynomial (lreman(a,b,n)+rreman(a,b,n))/2 STO
includes non-zero terms of integer degrees trapez(a,b,n)
from zero through order expanded in powers Simpson’s rule:
of (variable - point). 2*rreman(a,b,n)+trapez(a,b,n))/3 STO
For example taylor((sin(x), x, 5) gives simpson(a,b,n) *
x5/120 - x³/6 + x. And taylor((ln(x), x, 2, 1)
Now store your function in f(x) with STO
gives -(x-1)²/2 + x - 1.
f(x).
If the taylor series would include non-integer
For example to evaluate the midpoint Riemann
powers, the TI-89 must be forced to give the
sum for sin x from 0 to , with 100 intervals
expansion with a trick.
For example taylor(1/(x*(x-1)), x, 3) gives first key sin(x) STO f(x), then key
taylor(1/(x*(x-1)), x, 3, 0), but mreman(0, , 100). You can avoid having to
expand(taylor(1/(x-1)/x), x, 3), x) gives -x² - type mreman( in by keying 2nd VAR-LINK,
x - 1/x. scrolling down to mreman and keying
And taylor(e^(x)), x, 2) gives ENTER. mreman(0, , 100) gives 2. In exact
taylor(e^((x)), x, 2, 0), but taylor(e^(t), t, mode the TI-89 gives the exact value, in this
2)| t = (x) gives x/2 +x + 1. example, 4 cos (39/200) + 4 cos (37/200)
+ ... [less than 100 terms, because it goes
Riemann Sums through and simplifies them - very
impressive, but useless] after about 90
Riemann Sums are approximations to the seconds. So use auto mode.
definite integral. One sums x  ], Implicit differentiation
where x is the (uniform) width of the There is no implicit differentiation function, but
interval and f(zi) is the height of the ith you can create one. First move all of the
interval. For a sum with n terms over the terms of your equation to one side, i.e.
interval [a, b], x = (b-a)/n. The expression expression = 0.
for zi depends on whether one is doing the Define impdif (uu, xx, yy)
left, middle or right Riemann sum (RL, RM or = -d(uu,xx)/d(uu,yy), where uu is the
RR). expression to be implicitly differentiated,
Fortunately http://tifaq.calc.org/p2.htm#11.6 and d() is the derivative function. The
shows how to create functions to do these 3 variables xx and yy are used to avoid a
sums and also how to create functions to do circular definition error.
the Trapezoid and Simpson’s rules. (Check E.g. to implicitly differentiate x² = -y², key
out this awesome site!) Here’s how: impdif(x^2+y^2,x,y), which gives -x/y.
Type the following on the home screen, one at a This tip was also stolen from
time. http://tifaq.calc.org/p2.htm#11.6
RL:
* If you wish to also see these functions graphed
with the rectangle divisions drawn in,
programs for that are available from the
Houghton Mifflin website :
http://www.hmco.com/college/mathematics/b
erresford_rockett/appl_calc/gcprog.html.