---
title: "Second Moment Method"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Second Moment Method}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

The second moment method for project risk analysis is a quantitative technique used to estimate and manage the uncertainty in project outcomes. This method focuses on the first two moments of the probability distribution of project outcomes: the mean (or expected value) and the variance.

## Steps in the Second Moment Method

### 1. Identify Project Activities and Risks

List all the activities involved in the project; Identify the risks associated with each activity;  Estimate the mean (expected value) for each activity.

### 2. Determine the expected duration, cost, or other metrics for each activity considering the identified risks

The expected value is the average outcome, calculated as the weighted sum of all possible outcomes, where the weights are the probabilities of those outcomes.

### 3. Estimate Variance and Standard Deviation

Calculate the variance of each activity's duration, cost, or other metrics. Variance measures the spread of possible outcomes around the mean.

The standard deviation is the square root of the variance, providing a measure of uncertainty in the same units as the original metric.

### 4. Aggregate Mean and Variance

For the entire project, aggregate the means of all activities to get the project's overall expected duration, cost, etc.

Aggregate the variances to get the project's overall variance. If activities are independent, the total variance is the sum of the individual variances.

### 5. Assess Project Risk

Use the overall mean and variance to assess the project's risk. A larger variance indicates greater uncertainty and risk.

Calculate confidence intervals to understand the range within which the project outcomes are likely to fall. For example, assuming a normal distribution, approximately 95% of outcomes will fall within two standard deviations of the mean.

## Correlation

When accounting for correlation between tasks, the total variance of the project is calculated using both the individual variances and the covariances between each pair of activities.

## Formulation

### Expected Value (Mean):

For a given project activity \(i\), the expected value (mean) of the duration, cost, or any other metric \(X_i\) is:

\[ E(X_i) = \sum_{j} P_j \cdot X_{ij} \]

For the entire project, if there are \(n\) activities, the total expected value \(E(X)\) is:

\[ E(X) = \sum_{i=1}^{n} E(X_i) \]

### Variance:

The variance of a project activity \(i\) is:

\[ Var(X_i) = \sum_{j} P_j \cdot (X_{ij} - E(X_i))^2 \]

### Covariance:

The covariance between two activities \(X_i\) and \(X_j\) is:

\[ Cov(X_i, X_j) = \sum_{k} P_k \cdot (X_{ik} - E(X_i)) \cdot (X_{jk} - E(X_j)) \]

### Variance with Correlation:

When considering the correlation between activities, the total variance \(Var(X)\) of the project is:

\[ Var(X) = \sum_{i=1}^{n} Var(X_i) + 2 \sum_{i=1}^{n-1} \sum_{j=i+1}^{n} Cov(X_i, X_j) \]

### Correlation Coefficient:

The correlation coefficient \(\rho_{ij}\) between activities \(i\) and \(j\) is:

\[ \rho_{ij} = \frac{Cov(X_i, X_j)}{\sigma(X_i) \cdot \sigma(X_j)} \]

## Example 

First, load the package:
``` {r}
library(PRA)
```

Set the mean vector, variance vector, and correlation matrix for a toy project:
```{r}
mean <- c(10, 15, 20)
var <- c(4, 9, 16)
cor_mat <- matrix(c(
  1, 0.5, 0.3,
  0.5, 1, 0.4,
  0.3, 0.4, 1
), nrow = 3, byrow = TRUE)
```

Use the Second Moment Method to estimate the results for the project and print the results:
```{r results='asis'}
result <- smm(mean, var, cor_mat)
cat("Mean Total Cost is ", round(result$total_mean, 2))
cat("Variance around the Total Cost is ", round(result$total_var, 2))
```

## Benefits and Limitations

### Benefits

Simplicity: The method is relatively straightforward and easy to understand.

Quantitative Insight: Provides a numerical estimate of uncertainty and risk.

### Limitations

Normal Distribution Assumption: Often assumes that outcomes are normally distributed, which may not accurately represent all real-world scenarios.

Limited to First Two Moments: Ignores higher-order moments such as skewness and kurtosis, which can also be important in risk analysis.