A system is an intersection
Two equations in $x$ and $y$ can be typed exactly as printed — no rearranging into $y=\ldots$ first. Desmos draws each and the crossing point is the solution.
Solve $\;4x+3y=27\;$ and $\;2x-y=1$.
Typed as they stand, the lines cross at $(3,5)$. By hand: $y=2x-1$, so $4x+3(2x-1)=27$ gives $10x=30$, $x=3$ and $y=5$. $\boxed{(3,5)}$
Solve $\;3x+2y=16\;$ and $\;5x-4y=10$.
The crossing reads $(3.818,\,2.273)$. Those are rounded. Doubling the first equation gives $6x+4y=32$; adding the second gives $11x=42$, so $x=\tfrac{42}{11}$ and $y=\tfrac{16-3x}{2}=\tfrac{25}{11}$. Check: $\tfrac{42}{11}=3.8\overline{18}$ and $\tfrac{25}{11}=2.\overline{27}$, matching the display. $\boxed{\left(\tfrac{42}{11},\tfrac{25}{11}\right)}$
Với hệ hai phương trình bậc nhất, giao điểm trên màn hình luôn cho đúng vị trí nhưng không cho đúng giá trị khi nghiệm là phân số. Cách xử lý nhanh: nhìn số thập phân, đoán mẫu số nhỏ ($2,3,4,5,6,8,11$), gõ phân số đoán được vào một dòng mới và so chữ số. Mất $5$ giây, đổi lại đáp án chính xác cho câu tự điền.
Zero, one, or infinitely many
For two lines the picture and the algebra must agree, and the algebra is what the question usually asks about.
- $\dfrac{a_1}{a_2}\neq\dfrac{b_1}{b_2}$: they cross once — one solution.
- $\dfrac{a_1}{a_2}=\dfrac{b_1}{b_2}\neq\dfrac{c_1}{c_2}$: parallel and distinct — no solution.
- $\dfrac{a_1}{a_2}=\dfrac{b_1}{b_2}=\dfrac{c_1}{c_2}$: the same line — infinitely many solutions.
How many solutions does the system $\;y=x^2-4,\;y=3x-8\;$ have?
Substituting: $x^2-4=3x-8$, so $x^2-3x+4=0$ with discriminant $9-16=-7<0$: no real solution. The graph shows the line passing entirely below the parabola. $\boxed{0}$
Nhìn hai đường “gần trùng nhau” trên màn hình rồi kết luận vô số nghiệm. Trùng nhau và song song rất khó phân biệt bằng mắt khi hệ số gần bằng nhau. Muốn chắc: ẩn một đồ thị đi, nếu đường còn lại vẫn nằm đúng chỗ cũ thì hai đường trùng nhau; hoặc so ba tỉ lệ hệ số.
Tables and regressions
Press the plus button and choose table. Two columns appear, named $x_1$ and $y_1$, and every row you fill is plotted as a point.
A table on its own answers a whole family of questions: “which equation describes the relationship shown” becomes “type the four answer choices and see which curve goes through all the dots”.
A table lists $x=0,1,2,3$ against $y=7,\;11.2,\;15.4,\;19.6$. What is $y$ when $x=10$?
The differences are $4.2$ each time, so the model is $y=7+4.2x$, and
$y(10)=7+42=\boxed{49}$. In Desmos, enter the table and then a new line
y_1 ~ mx_1+b; the panel reports $m=4.2$, $b=7$, $R^2=1$.
Fitting a model
A regression line uses the tilde, not the equals sign. Typing y_1 ~ mx_1+b under a
table makes Desmos choose $m$ and $b$ to fit the points, and it reports both together with
$R^2$.
Linear: y_1 ~ mx_1+b
Quadratic: y_1 ~ ax_1^2+bx_1+c
Exponential: y_1 ~ a*b^x_1
$R^2=1$ means the model passes through every point exactly.
A population is $1200,\;1380,\;1587,\;1825.05$ at times $t=0,1,2,3$ years. Predict the population at $t=6$.
Consecutive ratios are all $1.15$, so $P(t)=1200(1.15)^t$ and $P(6)=1200(1.15)^6=1200(2.313\ldots)=2775.67\ldots\approx\boxed{2776}$. An exponential regression returns $a=1200$, $b=1.15$, $R^2=1$ — the same model, obtained without spotting the ratio.
A table lists $x=1,2,3,4,5$ against $y=3.2,\;4.1,\;5.3,\;5.9,\;7.0$. Using the line of best fit, what value is predicted at $x=8$?
Here $\bar x=3$ and $\bar y=5.1$. The slope is $\dfrac{\sum(x-\bar x)(y-\bar y)}{\sum(x-\bar x)^2}=\dfrac{9.4}{10}=0.94$, and the intercept is $5.1-0.94(3)=2.28$. So $\hat y=2.28+0.94(8)=\boxed{9.8}$. Desmos returns exactly these two numbers in one step; the arithmetic above is shown only so that the output is recognisable rather than magic.
Hai loại bảng số phải phân biệt. Bảng chính xác (mọi điểm nằm đúng trên một đường, $R^2=1$): dùng mô hình để tính giá trị bất kỳ, kể cả ngoài bảng. Bảng dữ liệu thật ($R^2<1$): chỉ được dùng từ “predicted” và không được nói giá trị dự đoán là giá trị thật. Đề hay bẫy đúng chỗ này bằng cách hỏi “theo đường khớp nhất, giá trị dự đoán là bao nhiêu” rồi đặt một phương án bằng đúng số liệu có sẵn trong bảng.
Choosing between four models
A large family of questions shows a table or a scatterplot and offers four candidate equations. Two routes work, and the second is almost always faster.
- Plot everything. Enter the table, then type all four choices and hide them one at a time with the coloured circles. Certain, but four expressions to type.
- Separate with one value. Pick the $x$ that splits the choices fastest — very often $x=0$, then $x=1$ — and evaluate mentally. Usually two choices die at once.
A table lists $x=0,1,2,3$ against $y=3,\;6,\;12,\;24$. Which equation fits?
- $y=3x+3$
- $y=3(2)^x$
- $y=3x^2$
- $y=2(3)^x$
At $x=0$: (A) gives $3$, (B) gives $3$, (C) gives $0$, (D) gives $2$. Choices (C) and (D) are gone after one substitution. At $x=2$: (A) gives $9$, (B) gives $12$. The table says $12$. $\boxed{\text{(B)}}$ — two substitutions, no typing.
A line of best fit for a scatterplot of years since $2010$ against membership has equation $\hat y=4.2x+180$. A question asks what $4.2$ represents.
The slope is the predicted change in $y$ for a one-unit increase in $x$, so membership is predicted to rise by about $\boxed{4.2\text{ members per year}}$. The intercept $180$ is the predicted membership in $2010$. Answer choices in this family almost always include “total members” (that is $\hat y$) and “$4.2$ years” (units swapped).
For the same fit, one data point is $(5,\,208)$. What is its residual?
The prediction is $\hat y=4.2(5)+180=201$, and the actual value is $208$, so the residual is $208-201=\boxed{7}$. A positive residual means the point sits above the line, so the model underestimates there.
Đảo thứ tự khi tính phần dư. Phần dư là thực tế trừ dự đoán, không phải ngược lại — đảo dấu là mất điểm dù số đúng. Dấu dương nghĩa là điểm nằm trên đường, mô hình dự đoán thiếu.
Sliders and parameter questions
Typing a letter other than $x$ and $y$ makes Desmos offer a slider for it. Dragging the slider redraws the graph in real time, which turns “for what value of $k$ …” into “drag until the picture is right, then read $k$”.
The slider is for finding the neighbourhood and for checking answer choices. The exact value still comes from algebra, because a slider moves in steps and stops on round numbers.
For what value of $k$ do the graphs of $y=x^2+k$ and $y=4x-1$ meet at exactly one point?
Setting them equal: $x^2+k=4x-1$, so $x^2-4x+(k+1)=0$. One solution means discriminant zero: $16-4(k+1)=0$, giving $k+1=4$ and $\boxed{k=3}$. Dragging a slider on $k$ shows the parabola rising until it just touches the line at $k=3$.
The system $\;3x+ky=12,\;6x+8y=5\;$ has no solution. What is $k$?
Parallel lines need $\dfrac{3}{6}=\dfrac{k}{8}$, so $k=4$. Then the first equation doubles to $6x+8y=24$, which contradicts $6x+8y=5$, so the lines really are distinct and there is indeed no solution. $\boxed{k=4}$
Kéo thanh trượt tới khi “trông giống” rồi ghi số hiển thị. Thanh trượt mặc định nhảy theo bước $0.001$ hoặc $0.1$, nên $k=3$ và $k=2.999$ cho hai hình gần như y hệt. Dùng thanh trượt để chọn giữa bốn phương án, còn câu tự điền thì phải giải bằng biệt thức hoặc tỉ lệ hệ số.
Lists and one-line statistics
A list is typed with square brackets, on a line of its own:
L=[4,7,7,9,12,15,15,18,40]
Once the list exists, mean(L), median(L), stdev(L),
total(L) and quartile(L,1) each return a number immediately.
For the data $4,7,7,9,12,15,15,18,40$, find the mean and the median, and say which is larger.
The total is $127$, so the mean is $\tfrac{127}{9}=14.11\ldots$; the fifth of nine ordered values is the median, $12$. The mean is larger. $\boxed{\text{mean }14.11,\ \text{median }12}$ — the single value $40$ drags the mean up and leaves the median untouched, which is exactly what “skewed to the right” means.
Sample $P$ is $10,12,14,16,18$ and sample $Q$ is $4,9,14,19,24$. Both have mean $14$. Which has the greater standard deviation, and what happens to each standard deviation if $5$ is added to every value?
$Q$ spreads from $4$ to $24$ while $P$ spreads from $10$ to $18$, so $Q$ has the greater standard deviation — $\texttt{stdev}$ returns about $3.16$ for $P$ and about $7.91$ for $Q$. Adding $5$ to every value slides both samples along the number line without changing any distance between values, so both standard deviations are unchanged while both means become $19$. $\boxed{Q\ \text{is more spread out; the standard deviations do not change}}$
Ba câu hỏi lặp đi lặp lại ở phần dữ liệu: bỏ một giá trị lớn ra khỏi mẫu thì trung bình đổi
nhiều hay trung vị đổi nhiều (trung bình); hai mẫu cùng trung bình thì mẫu nào độ lệch chuẩn
lớn hơn (mẫu trải rộng hơn); cộng thêm cùng một số vào mọi giá trị thì độ lệch chuẩn đổi
không (không đổi). Ba câu này trả lời được bằng lập luận nhanh hơn bấm máy — nhưng khi cần
số cụ thể thì stdev có sẵn.