21 KiB
Probability and Statistics ( BTech CSE )
- Ungrouped Data
- Grouped Data
- Relation between Mean, Median and Mode
- Relation between raw and central moments
- Skewness and Kurtosis
- Basic Probability
- Probability Function
- Proability Mass Function
- Moment Generating Function
- Binomial Distribution
- Poisson Distribution
- Exponential Distribution
- Normal Distribution
- Standard Normal Distribution
- Joint Probability Mass Function
- Useful equation
- Covariance in discrete data
- Regression
- Sampling
Ungrouped Data
Ungrouped data is data that has not been arranged in any way.So it is just a list of observations
\[ x_1, x_2, x_3, ... x_n \]
Mean
\[ \bar{x} = \frac{x_1 + x_2 + x_3 + ... + x_n}{n} \]
\[ \bar{x} = \frac{ \sum_{i = 1}^{n} x_i }{n} \]
Mode
The observation which occurs the highest number of time. So the x_i which has the highest count in the observation list.
Median
The median is the middle most observations. After ordering the n observations in observation list in either Ascending or Descending order (any order works). The median will be :
- n is even
\[ Median = \frac{ x_\frac{n}{2} + x_{(\frac{n}{2}+1)} }{2} \]
- n is odd
\[ Median = x_\frac{n+1}{2} \]
Variance and Standard Deviation
\[ Variance = \sigma^2 \] \[ Standard\ deviation = \sigma \]
\[ \sigma^2 = \frac{\sum_{i=1}^{n} (x_i - Mean)^2 }{n} \]
\[ \sigma^2 = \frac{\sum_{i=1}^n x_i^2}{n} - (Mean)^2 \]
Moments
About some constant A
\[ r^{th}\ moment = \frac{1}{n} \Sigma(x_i - A)^r \]
About Mean (Central Moment)
When A = Mean, then the moment is called central moment.
\[ \mu_r = \frac{1}{n} \Sigma(x_i - Mean)^r \]
About Zero (Raw Moment)
When A = 0, then the moment is called raw moment.
\[ \mu_r^{'} = \frac{1}{n} \Sigma x_i^r \]
Grouped Data
Data which is grouped based on the frequency at which it occurs. So if 9 appears 5 times in our observations, we group as x(observation) = 9 and f (frequency) = 5.
| x (observations) | f (frequency) |
| 2 | 5 |
| 1 | 3 |
| 4 | 5 |
| 8 | 9 |
If we store it in data way, i.e. the observations are of form 10-20, 20-30, 30-40 … then we will get $x_i$ by doing
\[ x_i = \frac{lower\ limit + upper\ limit}{2} \]
i.e,
$x_i$ for 20-30 will be $\frac{20 + 30}{2}$
So for data
| f (frequency) | |
| 0- 20 | 2 |
| 20-40 | 6 |
| 40-60 | 1 |
| 60-80 | 3 |
the $x_i$'s will become.
| f_i | x_i | |
| 0- 20 | 2 | 10 |
| 20-40 | 6 | 30 |
| 40-60 | 1 | 50 |
| 60-80 | 3 | 70 |
Mean
\[ \bar{x} = \frac{ \Sigma f_i x_i}{\Sigma f_i } \]
Mode
The modal class is the record with the row with the highest f_i
\[ Mode = l + (\frac{f_1 - f_0}{2f_1 - f_0 - f_2}) \times h \]
In the formula :
l → lower limit of modal class
f_1 → frequency(f_i) of the modal class
f_0 → frequency of the row preceding modal class
f_2 → frequency of the row after the modal class
h → size of class interval (upper limit - lower limit)
Median
The median for grouped data is calculated with the help of cumulative frequency. The cumulative frequency (cf_i) is given by:
\[ cf_i = f_1 + f_2 + f_3 + ... + f_i \]
The median class is the class whose cf_i is just greater than or is equal to $\frac{\Sigma f}{2}$
\[ Median = l + (\frac{(n/2) - cf}{f}) \times h \]
In the formula :
l → lower limit of the median class
h → size of class interval (upper limit - lower limit)
n → number of observations
cf → cumulative frequency of the median class
f → frequency of the median class
Variance and Standard Deviation
\[ Variance = \sigma^2 \] \[ Standard\ deviation = \sigma \]
\[ \sigma^2 = \frac{\sum_{i=1}^{n} f_i(x_i - Mean)^2 }{\Sigma f_i} \]
\[ \sigma^2 = \frac{\sum_{i=1}^n f_ix_i^2}{\Sigma f_i} - (Mean)^2 \]
Moments
About some constant A
\[ r^{th}\ moment = \frac{1}{\Sigma f_i} [\Sigma f_i (x_i - A)^r] \]
About Mean (Central Moment)
When A = Mean, then the moment is called central moment. \[ \mu_r = \frac{1}{\Sigma f_i} [\Sigma f_i (x_i - Mean)^r] \]
About Zero (Raw Moment)
When A = 0, then the moment is called raw moment. \[ \mu_r^{'} = \frac{1}{\Sigma f_i} [\Sigma f_i x_i^r] \]
Relation between Mean, Median and Mode
\[ 3Median = 2Mean + Mode \]
Relation between raw and central moments
\[ \mu_0 = \mu_0^{'} = 1 \] \[ \mu_1 = 0 \] \[ \mu_2 = \mu_2^{'} - \mu_1^{'2} \] \[ \mu_3 = \mu_3^{'} - 3\mu_1^{'}\mu_2^{'} + 2\mu_1^{'3} \] \[ \mu_4 = \mu_4^{'} - 4\mu_3^{'}\mu_1^{'} + 6\mu_2^{'}\mu_1^{'2} - 3\mu_1^{'4} \]
Skewness and Kurtosis
Skewness
- If Mean > Mode, then skewness is positive
- If Mean = Mode, then skewness is zero (graph is symmetric)
- If Mean < Mode, then skewness is zero
/Documents/ProbabilityAndStatistics/src/commit/270b93fd87e7fcce97630e1c8fb9d8281a8c3e40/skewness.PNG
Pearson's coefficient of skewness
The pearson's coefficient of skewness is denoted by SKP
\[ S_{KP} = \frac{Mean - Mode}{Standard\ Deviation} \]
- If SKP is zero then distribution is symmetrical
- If SKP is positive then distribution is positively skewed
- If SKP is negative then distribution is negatively skewed
Moment based coefficient of skewness
The moment based coefficient of skewness is denoted by β_1. The μ here is central moment.
\[ \beta_1 = \frac{\mu_3^2}{\mu_2^3} \]
The drawback of using β_1 as a coefficient of skewness is that it can only tell if distribution is symmetrical or not ,when $\beta_1 = 0$. It can't tell us the direction of skewness, i.e positive or negative.
- If β_1 is zero, then distribution is symmetrical
Karl Pearson's γ_1
To remove the drawback of the β_1 , we can derive Karl Pearson's γ_1
\[ \gamma_1 = \sqrt{\beta_1} \] \[ \gamma_1 = \frac{\mu_3}{\mu_2^{3/2}} \]
- If μ_3 is positive, the distribution has positive skewness
- If μ_3 is negative, the distribution has negative skewness
- If μ_3 is zero, the distribution is symmetrical
Kurtosis
Kurtosis is the measure of the peak and the curve and the "fatness" of the curve.
/Documents/ProbabilityAndStatistics/src/commit/270b93fd87e7fcce97630e1c8fb9d8281a8c3e40/kurtosis.PNG
The kurtosis is calculated using β_2
\[ \beta_2 = \frac{\mu_4}{\mu_2^2} \]
The value of β_2 tell's us about the type of curve
- Leptokurtic (High Peak) when β_2 > 3
- Mesokurtic (Normal Peak) when β_2 = 3
- Platykurtic (Low Peak) when β_2 < 3
Karl Pearson's γ_2
γ_2 is defined as:
\[ \gamma_2 = \beta_2 - 3 \]
- Leptokurtic when γ_2 > 0
- Mesokurtic when γ_2 = 0
- Platykurtic when γ_2 < 0
Basic Probability
Conditional Probability
If some event B has already occured, then the probability of the event A is:
\[ P(A \mid B) = \frac{P(A \cap B)}{P(B)} \]
$P(A \mid B)$ is read as A given B. So we are given that B has occured and this is probability of now A occuring.
Law of Total Probability
The law of total probability is used to find probability of some event A that has been partitioned into several different places/parts.
\[ P(A) = P(A|B_1)P(B_1) + P(A|B_2)P(B_2) + P(A|B_3)P(B_3) + ... + P(A|B_i)P(B_i) \] \[ P(A) = \Sigma P(A|B_i)P(B_i) \]
Example, Suppose we have 2 bags with marbles
- Bag 1 : 7 red marbles and 3 green marbles
- Bag 2 : 2 red marbles and 8 green marbles
Now we select one bag at random (i.e, the probability of choosing any of the two bags is equal so 0.5). If we draw a marble, what is the probability that it is a green marble?
Sol. The green marbles are in parts in bag 1 and bag 2.
Let G be the event of green marble.
Let B_1 be the event of choosing the bag 1
Let B-2 be the event of choosing the bag 2 \\
Then, $P(G|B_1) = \frac{3}{7 + 3}$ and $P(G|B_2) = \frac{8}{2 + 8}$ \\ Now, we can use the law of total probability to get
\[ P(G) = P(G|B_1)P(B_1) + P(G|B_2)P(B_2) \]
Example 2, Suppose a there are 3 forests in a park.
- Forest A occupies 50% of land and 20% plants in it are poisonous
- Forest B occupies 30% of land and 40% plants in it are poisonous
- Forest C occupies 20% of land and 70% plants in it are poisonous
What is the probability of a random plant from the park being poisonous.
Sol. Since probability is equal across whole area of the park. Event A is plant being from Forest A, Event B is plant being from Forest B and Event C is plant being from Forest C. If event P is plant being poisonous, then using law of total probability,
\[ P(P) = P(P|A)P(A) + P(P|B)P(B) + P(P|C)P(C) \]
And we know P(A) = 0.5, P(B) = 0.3 and P(C) = 0.2. Also P(P|A) = 0.20, P(P|B) = 0.40 and P(P|C) = 0.70
Some basic identities
- Probabilities follow law of inclusion and exclusion
\[ P(A \cup B) = P(A) + P(B) - P(A \cap B) \]
- DeMorgan's Theorem
\[ P(\overline{A \cap B }) = P(\overline{A} \cup \overline{B}) \] \[ P(\overline{A \cup B }) = P(\overline{A} \cap \overline{B}) \]
- Some other Identity
\[ P(\overline{A} \cap B) + P(A \cap B) = P(B) \] \[ P(A \cap \overline{B}) + P(A \cap B) = P(A) \]
Probability Function
It is a mathematical function that gives probability of occurance of different possible outcomes. We use variables to represent these possible outcomes called random variables. These are represented by capital letters. Example, $X$, $Y$, etc. We use these random variables as: \\ Suppose X is flipping two coins. \[ X = \{HH, HT, TT, TH\} \] We can represent it as, \[ X = \{0, 1, 2, 3\} \]
Now we can write a probability function $P(X=x)$ for flipping two coins as :
| $x$ | $P(X=x)$ |
| 0 | 0.25 |
| 1 | 0.25 |
| 2 | 0.25 |
| 3 | 0.25 |
Another example is throwing two dice and our random variable $X$ is sum of those two dice.
| $x$ | $P(X=x)$ |
| 2 | $1/36$ |
| 3 | $2/36$ |
| 4 | $3/36$ |
| 5 | $4/36$ |
| 6 | $5/36$ |
| 7 | $6/36$ |
| 8 | $5/36$ |
| 9 | $4/36$ |
| 10 | $3/36$ |
| 11 | $2/36$ |
| 12 | $1/36$ |
Types of probability functions (Continious and Discrete random variables)
Based on the range of the Random variables, probability function has two different names.
- For discrete random variables it is called Probability Distribution function.
- For continious random variables it is called Probability Density function.
Proability Mass Function
If we can get a function such that,
\[ f(x) = P(X=x) \]
then $f(x)$ is called a Probability Mass Function (PMF).
Properties of Probability Mass Function
Suppose a PMF
\[ f(x) = P(X=x) \]
Then,
For discrete variables
\[ \Sigma f(x) = 1 \] \[ E(X^n) = \Sigma x^n f(x) \]
For $E(X)$, the summation is over all possible values of x.
\[ Mean = E(X) = \Sigma x f(x) \] \[ Variance = E(X^2) - (E(X))^2 = \Sigma x^2 f(x) - ( \Sigma x f(x) )^2 \]
To get probabilities
\[ P(a \le X \le b) = \sum_{a}^{b} f(x) \] \[ P(a < X \le b) = (\sum_{a}^{b} f(x)) - f(a) \] \[ P(a \le X < b) = (\sum_{a}^{b} f(x)) - f(b) \]
Basically, we just add all $f(x)$ values from range of samples we need.
For continious variables
\[ \int_{-\infty}^{\infty} f(x) dx = 1 \] \[ E(X^n) = \int_{-\infty}^{\infty} x^n f(x) dx \]
We only consider integral from the possible values of x. Else we assume 0.
\[ Mean = E(X) = \int_{-\infty}^{\infty} x f(x) dx \] \[ Variance = E(X^2) - (E(X))^2 = \int_{-\infty}^{\infty} x^2 f(x) dx - ( \int_{-\infty}^{\infty} x f(x) dx )^2 \]
To get probability from a to b (inclusive and exclusive doesn't matter in continious).
\[ P(a < X < b) = \int_{a}^{b} f(x) dx \]
Some properties of mean and variance
- Mean
\[ E(aX) = aE(X) \] \[ E(a) = a \] \[ E(X + Y) = E(X) + E(Y) \]
- Variance
Variance is \[ V(X) = E(X^2) - (E(X))^2 \] Properties of variance are \[ V(aX) = a^2 V(X) \] \[ V(a) = 0 \]
Moment Generating Function
The moment generating function is given by
\[ M(t) = E(e^{tX}) \]
For discrete
\[ M(t) = \sum_{0}^{\infty} e^{tx} f(x) \]
For continious
\[ M(t) = \int_{-\infty}^{\infty} e^{tx} f(x) dx \]
Calculations of Moments (E(X)) using MGF
\[ E(X^n) = (\frac{d^n}{dt^n} M(t))_{t=0} \]
Binomial Distribution
The use of a binomial distribution is to calculate a known probability repeated n number of times, i.e, doing n number of trials. A binomial distribution deals with discrete random variables.
\[ X = \{ 0,1,2, .... n \} \]
where n is the number of trials.
\[ P(X=x) = \ ^nC_x\ (p)^x(q)^{n-x} \]
Here \[ n \rightarrow number\ of\ trials \] \[ x \rightarrow number\ of\ successes \] \[ p \rightarrow probability\ of\ success \] \[ q \rightarrow probability\ of\ failure \] \[ p = 1 - q \]
- Mean
\[ Mean = np \]
- Variance
\[ Variance = npq \]
- Moment Generating Function
\[ M(t) = (q + pe^t)^n \]
Additive Property of Binomial Distribution
For an independent variable $X$. The binomial distribution is represented as
\[ X ~ B(n,p) \] Here, \[ n \rightarrow number\ of\ trials \] \[ p \rightarrow probability\ of\ success \]
- Property
If given, \[ X_1 \sim B(n_1, p) \] \[ X_2 \sim B(n_2, p) \] Then, \[ X_1 + X_2 \sim B(n_1 + n_2, p) \]
- NOTE
If \[ X_1 \sim B(n_1, p_1) \] \[ X_2 \sim B(n_2, p_2) \] Then $X_1 + X_2$ is not a binomial distribution.
Using a binomial distribution
We can use binomial distribution to easily calculate probability of multiple trials, if probability of one trial is known. Example, the probability of a duplet (both dice have same number) when two dice are thrown is $\frac{6}{36}$.
Suppose now we want to know the probability of a 3 duplets if a pair of dice is thrown 5 times. So in this case :
\[ number\ of\ trials\ (n) = 5 \] \[ number\ of\ duplets\ we\ want\ probability\ for\ (x) = 3 \] \[ probability\ of\ duplet\ (p) = \frac{6}{36} \] \[ q = 1 - p = 1 - \frac{6}{36} \]
So using binomial distribution, \[ P(probability\ of\ 3\ duplets) = P(X=3) = \ ^5C_3 \left(\frac{6}{36}\right)^3 \left(\frac{30}{36}\right)^{5-3} \]
Poisson Distribution
A case of the binomial distribution where n is indefinitely large and p is very small and $\lambda = np$ is finite.
\[ P(X=x) = \frac{e^{-\lambda}\lambda^x}{x!}\ if\ x = 0, 1, 2 ..... \] \[ P(X=x) = 0\ otherwise \]
\[ \lambda = np \]
- Mean
\[ Mean = \lambda \]
- Variance
\[ Variance = \lambda \]
- Moment Generating Funtion
\[ M(t) = e^{\lambda\left(e^{t}-1\right)} \]
Additive property
If X_1, X_2, X_3..X_n follow poisson distribution with λ_1, λ_2, λ_3….λ_n
Then,
\[ X_1 + X_2 + X_3...+X_n \sim \lambda_1 + \lambda_2 + \lambda_3 + ...+ \lambda_n \]
Exponential Distribution
A continuous random distribution which has probability mass function
\[ f(x) = \lambda e^{-\lambda x}\ ,\ when\ x \ge 0 \] \[ f(x) = 0 \ ,\ otherwise \]
\[ where\ \lambda > 0 \]
- Mean
\[ Mean = \frac{1}{\lambda} \]
- Variance
\[ Variance = \frac{1}{\lambda^2} \]
- Moment Generating Function
\[ M(t) = \frac{\lambda}{\lambda - t} \]
Memory Less Property
\[ P[X > (s + t) \mid X > t] = P(X > s) \]
Normal Distribution
Suppose for a probability funtion with random variable X, having mean μ and variance σ^2.
We denote normal distribution using $X \sim N(\mu,\sigma)$
The probability mass funtion is
\[ f(x) = \frac{1}{\sigma\sqrt{2\pi}}\exp\left(-\frac{1}{2}\left(\frac{x-\mu}{\sigma}\right)^{2}\right) \]
\[ -\infty < x < \infty \] \[ -\infty < \mu < \infty \] \[ \sigma > 0 \] Here, $exp(x) = e^x$
- Moment Generating Funtion
\[ M(t) = exp\left( \mu t + \frac{\sigma^2 t^2}{2} \right) \]
Odd Moments
\[ E(X^{2n + 1}) = 0 \ , \ n = 0, 1, 2, ... \]
Even Moments
\[ E(X^{2n}) = 1.3.5....(2n-3)(2n-1) \sigma^{2n} \ , \ n = 0, 1, 2, ... \]
Properties
- In a normal distribution
\[ Mean = Mode = Median \]
- For normal distribution, mean deviation about mean is
\[ \sigma \sqrt{ \frac{2}{\pi} } \]
Additive property
Suppose for distributions X_1, X_2, X_3 … X_n with means μ_1 , μ_2 , μ_3 … μ_n and standard deviation σ_1^2 , σ_2^2 , σ_3^2 ….. σ_n^2 respectively. \\ Then X_1 + X_2 + X_3 will have mean ( μ_1 + μ_2 + μ_3 + … + μ_n ) and standard deviation (σ_1^2 + σ_2^2 + σ_3^2 + ….. + σ_n^2 )
- Additive Case
Given, \[ X_1 \sim N(\mu_1, \sigma_1) \] \[ X_2 \sim N(\mu_2, \sigma_2) \] Then, \[ a X_1 + b X_2 \sim N \left( a \mu_1 + b \mu_2, \sqrt{ a^2 \sigma_1^2 + b^2 \sigma_2^2} \right) \]
Standard Normal Distribution
The normal distribution with Mean 0 and Variance 1 is called the standard normal distribution.
\[ Z \sim N(0,1) \]
To calculate area under a given normal distribution, we can use the standard normal distribution. For that we need to calculate corresponding values in standard distribution from our given distribution. For that we have formula
\[ For\ X \sim N(\mu, \sigma) \] \[ z = \frac{x - \mu}{\sigma} \] \[ x \rightarrow value\ in\ our\ normal\ distribution \] \[ \mu \rightarrow mean\ of\ our\ distribution \] \[ \sigma \rightarrow standard\ deviation\ of\ our\ distribution \] \[ z \rightarrow corresponding\ value\ in\ standard\ normal\ distribution \]
Example,
Suppose for a normal distribution with X ∼ N(μ, σ) and we want to calculate probability P(a < X < b), then the ranges for same proability in the Z normal distribution will be,
\[ z_1 = \frac{a - \mu}{\sigma} \] \[ z_2 = \frac{b - \mu}{\sigma} \] Now the proability in Z distribution is, \[ P(z_1 < Z < z_2) \] \[ P( \frac{a - \mu}{\sigma} < Z < \frac{b - \mu}{\sigma} ) \]
So we need area under Z curve from a to b. \\ Then, we use the standard normal table to get the area.
- Note : The standard normal distribution is symmetric about the y axis. This fact can be used when calculating area under Z curve.
Joint Probability Mass Function
The joint probability mass distribution of two random variables X and Y is given by
\[ f(x,y) = P(X=x, Y=y) \]
- For discrete
\[ \Sigma_x \Sigma_y f(x,y) = 1 \]
- For continious
\[ \int_{-\infty}^{\infty} \int_{-\infty}^{\infty} f(x,y)\ dx\ dy = 1 \]
To get the probabilities,
\[ P(a \le X \le b, c \le Y \le d ) = \int_c^d \int_a^b f(x,y)\ dx\ dy\]
Marginal probability distribution (from joint PMF)
- For discrete
\[ P(X=x) = f(x) = \Sigma_y f(x,y) \] \[ P(Y=y) = f(y) = \Sigma_x f(x,y) \]
- For continious
\[ P(X=x) = f(x) = \int_{-\infty}^{\infty} f(x,y) dy \] \[ P(Y=y) = f(y) = \int_{-\infty}^{\infty} f(x,y) dx \]
Conditional Probability for Joint PMF
\[ P(X=x \mid Y=y) = f(x \mid y ) = \frac{ P(X=x, Y=y) }{ P(Y=y) } \] \[ P(X=x \mid Y=y) = f(x \mid y) = \frac{ f(x,y) }{ f(y) } \]
Independant Random Variables
The random variables X and Y are independant if, \[ f(x,y) = f(x) f(y) \]
Moment of Joint Variables
\[ E(X,Y) = E(XY) = \int_{-\infty}^\infty \int_{-\infty}^\infty xyf(x,y) dx\ dy \]
Covaraince
The covariance of two random variables X and Y is given by,
\[ cov(X,Y) = E(XY) - E(X)E(Y) \]
Properties of covariance
- If X and Y are independant
\[ cov(X,Y) = 0 \]
- If variance of some random variable X is written var(X), then
\[ cov(X+Y, X-Y) = var(X) - var(Y) \]
- General of previous case
\[ cov(aX + bY, cX + dY) = ac . var(X) + bd . var(Y) + (ad + bc) . cov(X,Y) \]
Variance of two random variables
\[ var(aX + bY) = a^2 . var(X) + b^2 . var(Y) + 2ab . cov(X,Y) \]
Correlation
The standard deviation of X is σ_X and standard deviation of Y is σ_Y. Then the correlation is given by,
\[ \gamma(X,Y) = \rho_{XY} = \frac{cov(X,Y)}{\sigma_X \sigma_Y } \]
here, ρXY lies between -1 and 1 \[ -1 \le \rho_{XY} \le 1 \]
Conditional moments
\[ E(X \mid Y) = \int_{-\infty}^{\infty} x f(x \mid y ) dx \ will\ be\ a\ function\ of\ y \]
Useful equation
\[ n! = \int_0^\infty x^n e^{-x} dx \]
Covariance in discrete data
Suppose for two sets of discrete data,
\[ X : x_1, x_2, x_3... x_n \] \[ Y : y_1, y_2, y_3... y_n \]
\[ cov(X,Y) = \frac{1}{n} \left( \sum_{i=1}^n x_i y_i \right) - [mean(x) . mean(y)] \]
\[ n \rightarrow number\ of\ items \]
Regression
Regression is a technique to relate a dependent variable to one or more independant variables.
Lines of regression
Both lines will pass through the point (mean(x) , mean(y))
y on x
Equation of line, \[ \frac{y - mean(y)}{x - mean(x)} = b_{yx} \] Where, \[ b_{yx} = \frac{cov(X,Y)}{var(Y)} \]
x on y
Equation of line, \[ \frac{x - mean(x)}{y - mean(y)} = b_{xy} \] Where, \[ b_{xy} = \frac{cov(X,Y)}{var(Y)} \]
bxy and byx are called regression coefficients.
- Note : if one of the regression coefficients is greater than 1, then the other must be less than 1.
Correlation
\[ \gamma(X,Y) = \rho_{XY} = \pm \sqrt{b_{xy} b_{yx}} \]
The sign of regression coefficients (bxy and byx) and the correlation coefficient is same.
Angle between lines of regression
\[ tan \theta = \left( \frac{ 1- \rho^2 }{ \rho } \frac{ \sigma_X . \sigma_Y }{ var(X) + var(Y) } \right) \]
Here σ is standard deviation.
- If $\rho = 0$ then $\theta = \frac{\pi}{2}$
- If $\rho = \pm 1$ then $\theta = 0$
TODO : Maybe an example here
Sampling
Notes not made for this currently, a pdf was provided by teacher as, ./sampling.pdf