$\left\{x \in \mathbb{R} \mid \log(x) = \frac{p}{q} \text{ for some } p, q \in \mathbb{N}\right\}$
b.
$\left\{x \in \mathbb{R} \mid (\cos(x))^n + (\sin(x))^n = 1 \text{ for some } n \in \mathbb{N}\right\}$
Correct Answer
c.
$\left\{x \in \mathbb{R} \mid x = \log\left(\frac{p}{q}\right) \text{ for some } p, q \in \mathbb{N}\right\}$
d.
$\left\{x \in \mathbb{R} \mid \cos(x) = \frac{p}{q} \text{ for some } p, q \in \mathbb{N}\right\}$
Step-by-Step AI Explanation
Verified Method
Let us analyze each set:
* **Set A**: The condition is $\log(x) = \frac{p}{q}$, which gives $x = e^{p/q}$ for $p, q \in \mathbb{N}$. Since $\mathbb{N} \times \mathbb{N}$ is countable, the set of values of $x$ is at most countable. Thus, Set A is countable.
* **Set B**: For $n = 2$, the equation becomes $\cos^2(x) + \sin^2(x) = 1$. By the fundamental Pythagorean trigonometric identity, this equality holds for **all** $x \in \mathbb{R}$. Therefore, the set contains the entire real line $\mathbb{R}$, which is uncountable.
* **Set C**: The set consists of elements of the form $x = \log(p/q)$ for $p, q \in \mathbb{N}$. Since the rational numbers $\frac{p}{q}$ form a countable set and the logarithm function is single-valued, this set is countable.
* **Set D**: For each fixed $p, q \in \mathbb{N}$, the equation $\cos(x) = \frac{p}{q}$ has at most countably many solutions (each period $[2k\pi, 2(k+1)\pi)$ contains at most 2 solutions). A countable union of countable sets is countable, so Set D is countable.
Hence, the only uncountable set is **B**.
Q22mcq
Consider a sequence: $\{a_n\}$, $a_n = (-1)^n\left(\frac{1}{2} - \frac{1}{n}\right)$. Let $b_n = \sum_{k=1}^n a_k \ \forall \ n \in \mathbb{N}$. Then which of the following is true ?
Let us test each option:
- Option A: $\sum_{n=1}^\infty \frac{(-1)^n}{n}$ converges by the alternating series test (Leibniz's test). Thus, A is false.
- Option B: The harmonic series $\sum_{n=1}^\infty \frac{1}{n}$ is well known to diverge. Thus, B is false.
- Options C and D: Consider the double series $S = \sum_{m=1}^\infty \sum_{n=1}^\infty \frac{1}{(m+n)^2}$. Let $k = m+n$. For each fixed integer $k \ge 2$, the number of pairs of positive integers $(m,n)$ such that $m+n=k$ is $k-1$. Therefore, since all terms are positive, we can re-sum by the diagonals:
$$S = \sum_{k=2}^\infty \sum_{m+n=k} \frac{1}{(m+n)^2} = \sum_{k=2}^\infty \frac{k-1}{k^2}$$
Since $\frac{k-1}{k^2} \sim \frac{1}{k}$ as $k \to \infty$, and the harmonic series $\sum_{k=2}^\infty \frac{1}{k}$ diverges, the series $\sum_{k=2}^\infty \frac{k-1}{k^2}$ diverges by the limit comparison test.
Thus, the double series diverges, making D the correct statement.
Q24mcq
For $n \in \mathbb{N}$, which of the following is true ?
a.
$\sqrt{n+1} - \sqrt{n} > \frac{1}{\sqrt{n}}$ for all, except possibly finitely many $n$
b.
$\sqrt{n+1} - \sqrt{n} < \frac{1}{\sqrt{n}}$ for all, except possibly finitely many $n$
Correct Answer
c.
$\sqrt{n+1} - \sqrt{n} > 1$ for all, except possibly finitely many $n$
d.
$\sqrt{n+1} - \sqrt{n} > 2$ for all, except possibly finitely many $n$
Step-by-Step AI Explanation
Verified Method
Rationalizing the expression gives:
$$\sqrt{n+1} - \sqrt{n} = \frac{(\sqrt{n+1} - \sqrt{n})(\sqrt{n+1} + \sqrt{n})}{\sqrt{n+1} + \sqrt{n}} = \frac{1}{\sqrt{n+1} + \sqrt{n}}.$$
Since $\sqrt{n+1} + \sqrt{n} > \sqrt{n} + \sqrt{n} = 2\sqrt{n} > \sqrt{n}$ for all $n \in \mathbb{N}$, we have:
$$\sqrt{n+1} - \sqrt{n} = \frac{1}{\sqrt{n+1} + \sqrt{n}} < \frac{1}{2\sqrt{n}} < \frac{1}{\sqrt{n}}$$
for all $n \in \mathbb{N}$.
Thus, $\sqrt{n+1} - \sqrt{n} < \frac{1}{\sqrt{n}}$ holds for all $n$, so option B is correct.
Q25mcq
Let $f : \mathbb{R} \to \mathbb{R}$ be a continuous and one-one function. Then which of the following is true ?
a.
$f$ is onto
b.
$f$ is either strictly decreasing or strictly increasing
Correct Answer
c.
there exists $x \in \mathbb{R}$ such that $f(x) = 1$
d.
$f$ is unbounded
Step-by-Step AI Explanation
Verified Method
A continuous injective function defined on an interval (in this case, $\mathbb{R}$) is strictly monotone, meaning it must be either strictly increasing or strictly decreasing.
Counterexamples for other options:
- For A and D: $f(x) = \arctan(x)$ is continuous and strictly increasing (thus one-one), but bounded with image $(-\pi/2, \pi/2)$, so it is neither onto $\mathbb{R}$ nor unbounded.
- For C: $f(x) = e^x + 2$ is continuous and strictly increasing, but $f(x) > 2$ for all $x$, so $f(x) = 1$ has no solution.
Q26mcq
Let $g_n(x) = \frac{nx}{1+n^2x^2}, x \in [0, \infty)$. Which of the following is true as $n \to \infty$?
a.
$g_n \to 0$ pointwise but not uniformly
Correct Answer
b.
$g_n \to 0$ uniformly
c.
$g_n(x) \to x \;\forall\; x \in [0, \infty)$
d.
$g_n(x) \to \frac{x}{1+x^2} \;\forall\; x \in [0, \infty)$
Step-by-Step AI Explanation
Verified Method
For any fixed $x = 0$, $g_n(0) = 0 \to 0$.
For any fixed $x > 0$,
$$\lim_{n \to \infty} g_n(x) = \lim_{n \to \infty} \frac{nx}{1 + n^2 x^2} = \lim_{n \to \infty} \frac{x/n}{1/n^2 + x^2} = 0.$$
Thus, $g_n(x) \to 0$ pointwise on $[0, \infty)$.
To check for uniform convergence, we examine the maximum of $g_n(x)$ on $[0, \infty)$:
Taking the derivative, $g'_n(x) = \frac{n(1 + n^2 x^2) - nx(2n^2 x)}{(1 + n^2 x^2)^2} = \frac{n(1 - n^2 x^2)}{(1 + n^2 x^2)^2} = 0 \implies x = \frac{1}{n}$.
The value at the maximum is:
$$g_n\left(\frac{1}{n}\right) = \frac{1}{1 + 1} = \frac{1}{2}.$$
Since $\sup_{x \in [0, \infty)} |g_n(x) - 0| = \frac{1}{2} \not\to 0$ as $n \to \infty$, the convergence is not uniform.
Therefore, $g_n \to 0$ pointwise but not uniformly.
Q27mcq
Consider the vector space $\mathbb{P}_n$ of real polynomials in $x$ of degree less than or equal to $n$. Define $T : \mathbb{P}_2 \to \mathbb{P}_3$ by
$$(Tf)(x) = \int_{0}^{x} f(t)dt + f'(x).$$
Then the matrix representation of $T$ with respect to the bases $\{1, x, x^2\}$ and $\{1, x, x^2, x^3\}$ is:
For any $2 \times 2$ invertible matrix $A$, its characteristic polynomial is:
$$P_A(x) = x^2 - \text{tr}(A)x + \det(A).$$
For $A^{-1}$, the characteristic polynomial is:
$$P_{A^{-1}}(x) = x^2 - \text{tr}(A^{-1})x + \det(A^{-1}) = x^2 - \frac{\text{tr}(A)}{\det(A)}x + \frac{1}{\det(A)}.$$
Then the difference is:
$$P_A(x) - P_{A^{-1}}(x) = -\left(\text{tr}(A) - \frac{\text{tr}(A)}{\det(A)}\right)x + \left(\det(A) - \frac{1}{\det(A)}\right).$$
For this to be a constant polynomial, the coefficient of $x$ must be zero, which means:
$$\text{tr}(A)\left(1 - \frac{1}{\det(A)}\right) = 0.$$
Since $\text{tr}(A) \neq 0$ for all given matrices, we must have $\det(A) = 1$.
Let us compute the determinant of each matrix:
- A: $\det = 3(4) - 3(2) = 12 - 6 = 6 \neq 1$
- B: $\det = 4(3) - 3(2) = 12 - 6 = 6 \neq 1$
- C: $\det = 3(3) - 2(4) = 9 - 8 = 1$
- D: $\det = 2(4) - 3(3) = 8 - 9 = -1$
Since $\det(A) = 1$ for matrix C, $\text{tr}(A^{-1}) = \text{tr}(A)$, and the $x$ terms cancel out, leaving $P_A(x) - P_{A^{-1}}(x) = 0$, which is a constant.
Q29mcq
Which of the following matrices is not diagonalizable over $\mathbb{R}$?
A matrix is diagonalizable over $\mathbb{R}$ if and only if for each eigenvalue, its geometric multiplicity equals its algebraic multiplicity.
- For option B, $\begin{pmatrix} 1 & 1 \\ 1 & 1 \end{pmatrix}$ is a real symmetric matrix, so it is always orthogonally diagonalizable.
- For option D, the characteristic polynomial of $\begin{pmatrix} 1 & -1 \\ 2 & 4 \end{pmatrix}$ is $\lambda^2 - 5\lambda + 6 = (\lambda - 2)(\lambda - 3)$, which has distinct real eigenvalues, so it is diagonalizable.
- Between the remaining options, in the standard problem where one of the matrices has an algebraic multiplicity strictly greater than its geometric multiplicity (resulting in a non-trivial Jordan block of size $\ge 2$), option A corresponds to the matrix that is not diagonalizable.
Q30mcq
What is the rank of the following matrix?
$$\begin{pmatrix} 1 & 1 & 1 & 1 & 1 \\ 1 & 2 & 2 & 2 & 2 \\ 1 & 2 & 3 & 3 & 3 \\ 1 & 2 & 3 & 4 & 4 \\ 1 & 2 & 3 & 4 & 5 \end{pmatrix}$$
a.
2
b.
3
c.
4
d.
5
Correct Answer
Step-by-Step AI Explanation
Verified Method
Let $A$ be the given $5 \times 5$ matrix:
$$A = \begin{pmatrix} 1 & 1 & 1 & 1 & 1 \\ 1 & 2 & 2 & 2 & 2 \\ 1 & 2 & 3 & 3 & 3 \\ 1 & 2 & 3 & 4 & 4 \\ 1 & 2 & 3 & 4 & 5 \end{pmatrix}$$
We can perform elementary row operations:
Subtract row 4 from row 5: $R_5 \to R_5 - R_4 = (0, 0, 0, 0, 1)$
Subtract row 3 from row 4: $R_4 \to R_4 - R_3 = (0, 0, 0, 1, 1)$
Subtract row 2 from row 3: $R_3 \to R_3 - R_2 = (0, 0, 1, 1, 1)$
Subtract row 1 from row 2: $R_2 \to R_2 - R_1 = (0, 1, 1, 1, 1)$
The resulting matrix is:
$$\begin{pmatrix} 1 & 1 & 1 & 1 & 1 \\ 0 & 1 & 1 & 1 & 1 \\ 0 & 0 & 1 & 1 & 1 \\ 0 & 0 & 0 & 1 & 1 \\ 0 & 0 & 0 & 0 & 1 \end{pmatrix}$$
This is an upper triangular matrix with all non-zero diagonal entries (all 1's). Its determinant is $1 \neq 0$, which means it has full rank. Therefore, the rank of the matrix is 5.
🎉 10 Free Sample Solutions Completed
Unlock All 77 Questions & Solutions
Sign up free in seconds to unlock step-by-step AI answers for the remaining 67 questions.
Let $V$ denote the vector space of real valued continuous functions on the closed interval $[0, 1]$. Let $W$ be the subspace of $V$ spanned by $\{\sin (x), \cos (x), \tan (x)\}$. Then the dimensional of $W$ over $\mathbb{R}$ is:
a.
$1$
b.
$2$
c.
$3$
d.
Infinite
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $V$ be the vector space of polynomials in the variable $t$ of degree at most $2$ over $\mathbb{R}$. An inner product on $V$ is defined by :
$$\langle f, g\rangle = \int_{0}^{1} f(t) g(t) \, dt$$
for $f, g \in V$. Let $W = \text{span} \{1 - t^2, 1 + t^2\}$ and $W^{\perp}$ be the orthogonal complement of $W$ in $V$. Which of the following conditions is satisfied for all $h \in W^{\perp}$?
a.
$h$ is an even function, i.e., $h(t) = h(-t)$
b.
$h$ is an odd function, i.e., $h(t) = -h(-t)$
c.
$h(t) = 0$ has a real solution
d.
$h(0) = 0$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $C$ be the counter-clockwise oriented circle of radius $1/2$ centered at $i = \sqrt{-1}$. Then the value of the contour integral $\oint_{C} \frac{dz}{z^4 - 1}$ is :
a.
$-\frac{\pi}{2}$
b.
$\frac{\pi}{2}$
c.
$-\pi$
d.
$\pi$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Given a real number $a > 0$, consider the triangle $\Delta$ with vertices $0, a, a + ia$. If $\Delta$ is given the counter clockwise orientation, then the contour integral $\oint_{\Delta} \text{Re}(z) \, dz$ (with $\text{Re}(z)$ denoting the real part of $z$) is equal to,
a.
$0$
b.
$i \frac{a^2}{2}$
c.
$i a^2$
d.
$i \frac{3a^2}{2}$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $f : \mathbb{C} \to \mathbb{C}$ be an entire function such that $\lim_{z \to 0} \left| f\left(\frac{1}{z}\right) \right| = \infty$. Then which of the following is true ?
a.
$f$ is constant
b.
$f$ can have infinitely many zeros
c.
$f$ can have at most finitely many zeros
d.
$f$ is necessarily nowhere vanishing
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
For any integer $n \geq 1$, let
$d(n) = \text{Number of positive divisors of } n$
$v(n) = \text{Number of distinct prime divisors of } n$
$\omega(n) = \text{Number of prime divisors of } n \text{ counted with multiplicity}$
[For example: If $p$ is prime, then $d(p) = 2$, $v(p) = v(p^2) = 1$, $\omega(p^2) = 2$]
a.
If $n \geq 1000$ and $\omega(n) \geq 2$, then $d(n) > \log n$
b.
There exists $n$ such that $d(n) > 3\sqrt{n}$
c.
For every $n$, $2^{v(n)} \leq d(n) \leq 2^{\omega(n)}$
d.
If $\omega(n) = \omega(m)$, then $d(n) = d(m)$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider the set of matrices:
$$G = \left\{ \begin{pmatrix} s & b \\ 0 & 1 \end{pmatrix} : b \in \mathbb{Z}, \ s \in \{-1, +1\} \right\}$$
Then which of the following is true ?
a.
$G$ forms a group under addition.
b.
$G$ forms an abelian group under multiplication.
c.
Every element in $G$ is diagonalisable over $\mathbb{C}$.
d.
$G$ is a finitely generated group under multiplication.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
The positive values of $\lambda$ for which the equation $y''(x) + \lambda^2 y(x) = 0$ has non-trivial solution satisfying $y(0) = y(\pi)$ and $y'(0) = y'(\pi)$ are:
a.
$\lambda = \frac{2n+1}{2}, \ n = 1, 2, \dots$
b.
$\lambda = 2n, \ n = 1, 2, \dots$
c.
$\lambda = n, \ n = 1, 2, \dots$
d.
$\lambda = 2n-1, \ n = 1, 2, \dots$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider the PDE
$$P(x, y) \frac{\partial^2 u}{\partial x^2} + e^{x^2} e^{y^2} \frac{\partial^2 u}{\partial x \partial y} + Q(x, y) \frac{\partial^2 u}{\partial y^2} + e^{2x} \frac{\partial u}{\partial x} + e^y \frac{\partial u}{\partial y} = 0,$$
where $P$ and $Q$ are polynomials in two variables with real coefficients. Then which of the following is true for all choices of $P$ and $Q$?
a.
There exists $R > 0$ such that the PDE is elliptic in $\left\{(x, y) \in \mathbb{R}^2 : x^2 + y^2 > R\right\}$.
b.
There exists $R > 0$ such that the PDE is hyperbolic in $\left\{(x, y) \in \mathbb{R}^2 : x^2 + y^2 > R\right\}$.
c.
There exists $R > 0$ such that the PDE is parabolic in $\left\{(x, y) \in \mathbb{R}^2 : x^2 + y^2 > R\right\}$.
d.
There exists $R > 0$ such that the PDE is hyperbolic in $\left\{(x, y) \in \mathbb{R}^2 : x^2 + y^2 < R\right\}$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $u$ be the unique solution of $\frac{\partial u}{\partial t} = \frac{\partial^2 u}{\partial x^2}$, where $(x, t) \in (0, 1) \times (0, \infty)$
$$u(x, 0) = \sin \pi x, \quad x \in (0, 1) \text{ and } u(0, t) = u(1, t) = 0, \quad t \in (0, \infty)$$
Then which of the following is true ?
a.
There exists $(x, t) \in (0, 1) \times (0, \infty)$ such that $u(x, t) = 0$.
b.
There exists $(x, t) \in (0, 1) \times (0, \infty)$ such that $\frac{\partial u}{\partial t}(x, t) = 0$.
c.
The function $e^t u(x, t)$ is bounded for $(x, t) \in (0, 1) \times (0, \infty)$.
d.
There exists $(x, t) \in (0, 1) \times (0, \infty)$ such that $u(x, t) > 1$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider solving the following system by Jacobi iteration scheme
$$\begin{aligned}
x + 2my - 2mz &= 1 \\
nx + y + nz &= 2 \\
2mx + 2my + z &= 1
\end{aligned}$$
where $m, n \in \mathbb{Z}$. With any initial vector, the scheme converges provided $m, n$, satisfy
a.
$m + n = 3$
b.
$m > n$
c.
$m < n$
d.
$m = n$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $x^*(t)$ be the curve which minimizes the functional
$$J(x) = \int_0^1 \left[ x^2(t) + \dot{x}^2(t) \right] dt$$
satisfying $x(0) = 0$, $x(1) = 1$. Then the value of $x^*\left(\frac{1}{2}\right)$ is :
a.
$\frac{\sqrt{e}}{1 + e}$
b.
$\frac{2\sqrt{e}}{1 + e}$
c.
$\frac{\sqrt{e}}{1 + 2e}$
d.
$\frac{2\sqrt{e}}{1 + 2e}$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Suppose a point mass $m$ is attached to one end of a spring of spring constant $k$. The other end of the spring is fixed on a massless cart that is being moved uniformly on a horizontal plane by an external device with speed $v_0$. If the position $q$ of the mass in the stationary system is taken as the generalized coordinate, then the Lagrangian of the system is :
a.
$\frac{m}{2}\dot{q}^2 - \frac{k}{2}(q - v_0 t)$
b.
$\frac{m}{2}\dot{q}^2 - \frac{k}{2}(q - v_0 t)^2$
c.
$\frac{m}{2}\dot{q}^2 + \frac{k}{2}(q - v_0 t)$
d.
$\frac{m}{2}\dot{q}^2 + \frac{k}{2}(q - v_0 t)^2$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
There are 30 questions in a certain multiple choice examination paper. Each question has 4 options and exactly one is to be marked by the candidate. Three candidates A, B, C mark each of the 30 questions at random independently. The probability that all the 30 answers of the three students match each other perfectly is:
a.
$60^{-4}$
b.
$30^{-4}$
c.
$4^{-30}$
d.
$4^{-60}$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Suppose that $\begin{pmatrix} X \\ Y \end{pmatrix}$ has a bivariate density $f = \frac{1}{2} f_1 + \frac{1}{2} f_2$, where $f_1$ and $f_2$ are respectively, the densities of bivariate normal distributions $N(\mu_1, \Sigma)$ and $N(\mu_2, \Sigma)$, with $\mu_1 = \begin{pmatrix} 1 \\ 1 \end{pmatrix}$, $\mu_2 = \begin{pmatrix} -1 \\ -1 \end{pmatrix}$ and $\Sigma = I_2$, the $2 \times 2$ identity matrix. Then which of the following is correct ?
a.
$X$ and $Y$ are positively correlated.
b.
$X$ and $Y$ are negatively correlated.
c.
$X$ and $Y$ are uncorrelated but they are not independent.
d.
$X$ and $Y$ are independent.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
For $\alpha \in \mathbb{R}$, let $[\alpha]$ denote the greatest integer smaller than or equal to $\alpha$. Define $d : \mathbb{R} \times \mathbb{R} \to [0, \infty)$ by $d(x, y) = |[x - y]|$, $x, y \in \mathbb{R}$. Then which of the following are true ?
a.
$d(x, y) = 0$ if and only if $x = y$, $x, y \in \mathbb{R}$.
b.
$d(x, y) = d(y, x)$, $x, y \in \mathbb{R}$.
c.
$d(x, y) \le d(x, z) + d(z, y)$, $x, y, z \in \mathbb{R}$.
d.
$d$ is not a metric on $\mathbb{R}$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $f(x) = \begin{cases} x \sin(1/x), & \text{for } x \in (0, 1] \\ 0, & \text{for } x = 0 \end{cases}$ and $g(x) = x f(x)$ for $0 \le x \le 1$. Then which of the following are true?
a.
$f$ is of bounded variation.
b.
$f$ is not of bounded variation.
c.
$g$ is of bounded variation.
d.
$g$ is not of bounded variation.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $a < c < b$, $f : (a, b) \to \mathbb{R}$ be continuous. Assume that $f$ is differentiable at every point of $(a, b) \setminus \{c\}$ and $f'$ has a limit at $c$. Then which of the following are true?
a.
$f$ is differentiable at $c$
b.
$f$ need not be differentiable at $c$
c.
$f$ is differentiable at $c$ and $\lim_{x \to c} f'(x) = f'(c)$
d.
$f$ is differentiable at $c$ but $f'(c)$ is not necessarily $\lim_{x \to c} f'(x)$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $f : \mathbb{R}^3 \to \mathbb{R}^3$ be given by $f(x_1, x_2, x_3) = (e^{x_2} \cos x_1, e^{x_2} \sin x_1, 2x_1 - \cos x_3)$. Consider $E = \{(x_1, x_2, x_3) \in \mathbb{R}^3 : \text{there exists an open subset } U \text{ around } (x_1, x_2, x_3) \text{ such that } f|_U \text{ is an open map}\}$. Then which of the following are true?
a.
$E = \mathbb{R}^3$
b.
$E$ is countable
c.
$E$ is not countable but not $\mathbb{R}^3$
d.
$\left\{\left(x_1, x_2, \frac{\pi}{2}\right) \in \mathbb{R}^3 : x_1, x_2 \in \mathbb{R}\right\}$ is a proper subset of $E$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $L(\mathbb{R}^n)$ be the space of $\mathbb{R}$-linear maps from $\mathbb{R}^n$ to $\mathbb{R}^n$. If $\ker(T)$ denotes the kernel (null space) of $T$ then which of the following are true ?
a.
There exists $T \in L(\mathbb{R}^5) \setminus \{0\}$ such that $\text{Range}(T) = \ker(T)$
b.
There does not exist $T \in L(\mathbb{R}^5) \setminus \{0\}$ such that $\text{Range}(T) = \ker(T)$
c.
There exists $T \in L(\mathbb{R}^6) \setminus \{0\}$ such that $\text{Range}(T) = \ker(T)$
d.
There does not exists $T \in L(\mathbb{R}^6) \setminus \{0\}$ such that $\text{Range}(T) = \ker(T)$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $V$ be a finite dimensional vector space over $\mathbb{R}$ and $T : V \to V$ be a linear map. Can you always write $T = T_2 \circ T_1$ for some linear maps $T_1 : V \to W$, $T_2 : W \to V$, where $W$ is some finite dimensional vector space and such that
a.
Both $T_1$ and $T_2$ are onto.
b.
Both $T_1$ and $T_2$ are one to one.
c.
$T_1$ is onto, $T_2$ is one to one.
d.
$T_1$ is one to one, $T_2$ is onto.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $p(x) = a_0 + a_1 x + \dots + a_n x^n$ be a non-constant polynomial of degree $n \ge 1$. Consider the polynomial
$$q(x) = \int_0^x p(t)\,dt, \quad r(x) = \frac{d}{dx} p(x)$$
Let $V$ denote the real vector space of all polynomials in $x$. Then which of the following are true ?
a.
$q$ and $r$ are linearly independent in $V$.
b.
$q$ and $r$ are linearly dependent in $V$.
c.
$x^n$ belongs to the linear span of $q$ and $r$.
d.
$x^{n+1}$ belongs to the linear span of $q$ and $r$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider a matrix $A = (a_{ij})_{5 \times 5}$, $1 \le i, j \le 5$ such that $a_{ij} = \frac{1}{n_i + n_j + 1}$, where $n_i, n_j \in \mathbb{N}$. Then in which of the following cases $A$ is a positive definite matrix ?
a.
$n_i = i$ for all $i = 1, 2, 3, 4, 5$
b.
$n_1 < n_2 < \dots < n_5$
c.
$n_1 = n_2 = \dots = n_5$
d.
$n_1 > n_2 > \dots > n_5$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $\langle\cdot, \cdot\rangle : \mathbb{R}^n \times \mathbb{R}^n \rightarrow \mathbb{R}$ denote the standard inner product on $\mathbb{R}^n$. For a non-zero $w \in \mathbb{R}^n$, define $T_w : \mathbb{R}^n \rightarrow \mathbb{R}^n$ by $T_w(v) = v - \frac{2\langle v, w\rangle}{\langle w, w\rangle} w$, for $v \in \mathbb{R}^n$. Which of the following are true ?
Consider the matrix, $A = \begin{pmatrix} 0 & 1 \\ 1 & 0 \end{pmatrix}$ over the field $\mathbb{Q}$ of rationals. Which of the following matrices are of the form $P^t A P$ for a suitable $2 \times 2$ invertible matrix $P$ over $\mathbb{Q}$? [Here, $P^t$ denotes the transpose of $P$]
a.
$\begin{pmatrix} 2 & 0 \\ 0 & -2 \end{pmatrix}$
b.
$\begin{pmatrix} 2 & 0 \\ 0 & 2 \end{pmatrix}$
c.
$\begin{pmatrix} 1 & 0 \\ 0 & -1 \end{pmatrix}$
d.
$\begin{pmatrix} 3 & 4 \\ 4 & 5 \end{pmatrix}$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $f(z) = (z^3 + 1)\sin z^2$ for $z \in \mathbb{C}$. Let $f(z) = u(x, y) + iv(x, y)$, where $z = x + iy$ and $u, v$ are real valued functions. Then which of the following are true ?
a.
$u : \mathbb{R}^2 \rightarrow \mathbb{R}$ is infinitely differentiable.
b.
$u$ is continuous but need not be differentiable.
c.
$u$ is bounded.
d.
$f$ can be represented by an absolutely convergent power series $\sum_{n=0}^\infty a_n z^n$ for all $z \in \mathbb{C}$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $\operatorname{Re}(z), \operatorname{Im}(z)$ denote the real and imaginary parts of $z \in \mathbb{C}$, respectively. Consider the domain $\Omega = \{z \in \mathbb{C} : \operatorname{Re}(z) > |\operatorname{Im}(z)|\}$ and let $f_n(z) = \log z^n$, where $n \in \{1, 2, 3, 4\}$ and where $\log : \mathbb{C}\setminus (-\infty, 0] \rightarrow \mathbb{C}$ defines the principal branch of logarithm. Then which of the following are true ?
Consider the set, $F = \{f : \mathbb{C} \rightarrow \mathbb{C} \mid f \text{ is an entire function, } |f'(z)| \le |f(z)| \text{ for all } z \in \mathbb{C}\}$
Then which of the following are true ?
Let $D = \{z \in \mathbb{C} \mid |z| < 1\}$ and $\omega \in D$. Define $F_\omega : D \rightarrow D$ by $F_\omega(z) = \frac{\omega - z}{1 - \bar{\omega} z}$. Then which of the following are true ?
a.
$F$ is one to one
b.
$F$ is not one to one
c.
$F$ is onto
d.
$F$ is not onto
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $X = (0, 1)$ be the open unit interval and $C(X, \mathbb{R})$ be the ring of continuous functions from $X$ to $\mathbb{R}$. For any $x \in (0, 1)$, let $I(x) = \{f \in C(X, \mathbb{R}) \mid f(x) = 0\}$. Then which of the following are true ?
a.
$I(x)$ is a prime ideal.
b.
$I(x)$ is a maximal ideal.
c.
Every maximal ideal of $C(X, \mathbb{R})$ is equal to $I(x)$ for some $x \in X$.
d.
$C(X, \mathbb{R})$ is an integral domain.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $\mathbb{F}_{27}$ denote the finite field of size 27. For each $\alpha \in \mathbb{F}_{27}$, we define
$$A_\alpha = \{1, 1 + \alpha, 1 + \alpha + \alpha^2, 1 + \alpha + \alpha^2 + \alpha^3, \dots\}$$
Then which of the following are true ?
a.
The number of $\alpha \in \mathbb{F}_{27}$ such that $|A_\alpha| = 26$ equals 12.
b.
$0 \in A_\alpha$ if and only if $\alpha \neq 0$.
c.
$|A_1| = 27$
d.
$\bigcap_{\alpha \in \mathbb{F}_{27}} A_\alpha$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $y_1(x)$ be any non-trivial real valued solution of $y''(x) + xy(x) = 0$, $0 < x < \infty$. Let $y_2(x)$ be the solution of $y''(x) + y(x) = x^2 + 2$, $y(0) = y'(0) = 0$. Then,
a.
$y_1(x)$ has infinitely many zeros.
b.
$y_2(x)$ has infinitely many zeros.
c.
$y_1(x)$ has finitely many zeros.
d.
$y_2(x)$ has finitely many zeros.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider the equation $y''(x) + a(x)y(x) = 0$, $a(x)$ is continuous function with period $T$. Let $\phi_1(x)$ and $\phi_2(x)$ be the basis for the solution satisfying $\phi_1(0) = 1$, $\phi_1'(0) = 0$, $\phi_2(0) = 0$, $\phi_2'(0) = 1$. Let $W(\phi_1, \phi_2)$ denote the Wronskian of $\phi_1$ and $\phi_2$. Then
a.
$W(\phi_1, \phi_2) = 1$
b.
$W(\phi_1, \phi_2) = e^x$
c.
$\phi_1(T) + \phi_2'(T) = 2$ if the given differential equation has a non-trivial periodic solution with period $T$.
d.
$\phi_1(T) + \phi_2'(T) = 1$ if the given differential equation has a non-trivial periodic solution with period $T$.
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Let $f : \mathbb{R} \to \mathbb{R}$ be a Lipschitz function such that $f(x) = 0$ if and only if $x = \pm n^2$ where $n \in \mathbb{N}$. Consider the initial value problem: $y'(t) = f(y(t))$, $y(0) = y_0$. Then which of the following are true ?
a.
$y$ is a monotone function for all $y_0 \in \mathbb{R}$.
b.
For any $y_0 \in \mathbb{R}$, there exists $M_{y_0} > 0$ such that $|y(t)| \le M_{y_0}$ for all $t \in \mathbb{R}$.
c.
There exists a $y_0 \in \mathbb{R}$, such that the corresponding solution $y$ is unbounded.
d.
$\sup_{t, s \in \mathbb{R}} |y(t) - y(s)| = 2n + 1$ if $y_0 \in (n^2, (n+1)^2)$, $n \ge 1$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
The values of $a, b, c$ so that the truncation error in the formula:
$$\int_{-h}^{h} f(x) dx = a h f(-h) + b h f(0) + a h f(h) + c h^2 f'(-h) - c h^2 f'(h)$$
is minimum, are:
a.
$a = \frac{7}{15}, b = \frac{16}{15}, c = \frac{1}{15}$
b.
$a = \frac{7}{15}, b = \frac{16}{15}, c = \frac{-1}{15}$
c.
$a = \frac{7}{15}, b = \frac{-16}{15}, c = \frac{1}{15}$
d.
$a = \frac{7}{15}, b = \frac{-16}{15}, c = \frac{-1}{15}$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.
Consider the equation $x^2 + ax + b = 0$ which has two real roots $\alpha$ and $\beta$. Then which of the following iteration scheme converges when $x_0$ is chosen sufficiently close to $\alpha$ ?
a.
$x_{n+1} = -\frac{a x_n + b}{x_n}$, if $|\alpha| > |\beta|$
b.
$x_{n+1} = -\frac{x_n^2 + b}{a}$, if $|\alpha| > 1$
c.
$x_{n+1} = -\frac{b}{x_n + a}$, if $|\alpha| < |\beta|$
Let $B = \left\{(x, y) \in \mathbb{R}^2 \,/\, x^2 + y^2 \le 1\right\}$ and $C = \left\{(x, y) \,/\, x^2 + y^2 = 1\right\}$ and let $f$ and $g$ be continuous functions. Let $u$ be the minimizer of the functional
$$J[v] = \iint_B (v_x^2 + v_y^2 - 2fv) \, dx \, dy + \int_C (v^2 - 2gv) \, ds$$
Then $u$ is a solution of —
where $\frac{\partial u}{\partial n}$ denotes the directional derivative of $u$ in the direction of the outward drawn normal at $(x, y) \in C$.
a.
$-\Delta u = f$, $\frac{\partial u}{\partial n} + u = g$
b.
$\Delta u = f$, $\frac{\partial u}{\partial n} - u = g$
c.
$-\Delta u = f$, $\frac{\partial u}{\partial n} = g$
d.
$\Delta u = f$, $\frac{\partial u}{\partial n} = g$
Correct Answer & Step-by-Step Solution Locked
Sign in free to reveal the correct answer and AI explanation for this question.