You are on page 1of 51

Networks: Chapter 23 – 24

  

From this graph→ The points A, B, C, D, E and F are called vertices or The lines connecting the vertices are called edges. set of elements called vertices and a set of elements edge joins two vertices. The vertices of A and B of a graph are adjacent joined by an edge. B 1 0 1 0 0 1 C 0 1 0 1 1 0 D 0 0 1 0 0 1 E 1 0 1 0 0 0 F 1 1 0 1 0 0

nodes of the graph. A has played B, F and E A consists a A B graph has played C, Fof and called edges. Each C has played B, D and E D has played F and C vertices if they are C E has played A and F has played B, A and D

A A B C D E F 0 1 0 0 1 1

The graph can then be represented by a table or matrix. A 1 is used to denote that there is one edge connecting the two vertices and a 0 indicates that there is no edge. The adjacency matrix A of the graph is an n x n matrix in which for example the entry in row C and column F is the number of edges joining vertices C and F. The graph is said to have multiple edges if there is more than 1 edge connecting any two vertices. When a vertice links itself with an edge, it is said to be a loop.

A biparite graph is a graph whose set of vertices can be split into two subsets X and Y in such a way that each edge of the graph joins a vertex in X and a vertex in Y. The degree of a vertex is written as deg(X) and is the number of edges incident with it. A simple graph is a graph with no loops or multiple edges. For a simple graph the sum of the degrees of the graph is equal to twice the number of edges of the graph – the sum of the degrees of a simple graph is even. An isolated vertex is a vertex which has no edges incident to it. A graph is said to be degenerate if all its vertices are isolated. A path is a sequence of edges that does not begin and finish at the same vertice A circuit is a sequence of edges linking successive vertices that starts and finishes at the same vertex. A graph is said to be connected if there is a path between each pair of vertices A subgraph of a graph consists of selected edges and vertices of the graph with the same links as the original graphs and with the selected vertices includign all the endpoints of the selected edges.

A graph that can be drawn in such a way that no two edges meet (or have common points), except at the vertices where they are both incident, is called a planar graph. Complete graphs cannot be planar. When counting the faces in a graph, a face must be counted on the outside of the graph, an infinite face. There is a relationship between the number of vertices, v, edges, e, and faces, f, in a connected planar graph: This is called Euler’s formula, which sates: A complete graph is a graph will edges connecting all pairs of vertices The complete graph with n vertices is denoted by Kn Complete graphs with 1-4 vertices are planar, whilst graphs with > 5 vertices are not planar The complete graph with n vertices, Kn has edges Complete graphs can be used to represent round robin competitions, in which each team plays each other once. A path that includes every edge just once is called an Euler path An Euler circuit is an Euler path that starts and finishes at the same vertex A useful way of identifying Euler circuits is to look for a connected graph where all vertices have an even degree A connected graph has an Euler path starting at vertex A and finishing at vertex B if A and B are the only odd vertices An Euler path will have exactly two vertices of odd degree, not 3, not 1, only two. The rest are of even degree. A Hamilton path is a path through a graph that passes through each vertex exactly once A Hamilton circuit is a Hamilton path which starts and finishes at the same vertex Hamilton graphs are defined in terms of vertices and Euler graphs are defined in terms of edges. Hamilton circuits cannot be identified by any specific condition like a Euler circuit/path – it is just trial and error A weighted graph is one where a number is associated with each edge – these numbers are called weights A tree is a connected graph that contains no circuits – a tree has no multiple edges The simplest tree is a single vertex – a tree with n vertices has n – 1 edges A spanning tree is a subgraph that contains all the vertices of the original graphs and is a tree A minimum spanning tree for a weighted graph is a spanning tree for which the sum of the weights of the edges is as small as possible Prim’s algorithm is a set of rules to determine a minimum spanning tree for a graph: 1. Choose a vertex and connect it to a second vertices chosen so that the weight of the edge is as small as possible 2. In each step thereafter, take the edge with the lowest weight, producing a tree with the edges already selected 3. Repeat until all the vertices are connected and then stop A graph where direction is indicated for every edge is called a directed graph – this is often abbreviated to digraph A network is a connected directed graph with no loops This information can again be represented with an adjacency matrix [read across horizontally to see who is victorious] One point in a network is said to be reachable from another different point in a directed graph if a path exists between the two points – in this situation, each edge and vertex can only be traversed once The matrix for reachability is written the same as an adjacency matrix, but is read vertically. = number of one step paths, = number of two step paths, = number of three step paths, We combine the information from the individual matrices into a single reachability matrix, R, which will show all of the possible paths in the directed graph (by combining the information in the 3 matrices by adding corresponding elements) Using the matrix, we can calculate the total reachability of each of the vertices by adding each of the columns A vertex in a directed graph is said to be dominant if a path exists between that vertex and another but not vice versa = number of one step paths, = number of two step paths, D Using the matrix, we can calculate the total dominance of each of the vertices by adding each of the rows A flow is the quantity of material that can move along a given channel (eg. A highway traffic flow) A cut is a method for determining the maximum flow for a network

A cut is defined as a collection of edges that, if removed from the directed graph, produces a zero flow between the start S and the terminal T The capacity of a cut is the sum of the capacities of the edges directed from S to T that the cut passes through Edges that cannot contribute to the flow of a network in the required direction are counted as zero Maximum flow capacity of any cut Maximum flow capacity of the minimum cut This can be strengthened to: Maximum flow = capacity of minimum cut The Mr. Ryan Method can also be used to determine the maximum flow of a network By deducting the lowest weight along any particular path from S to T from all weights along that path and noting the weight deduced, this process can be repeated until there is no longer all paths stretching from S to T have a “0” weight throughout their duration and consequently, there are not paths from start to finish – by adding the weights deducted throughout this process the maximum flow can be determined A critical activity is any task that, if delayed, will also delay the entire project The weight of the critical path is the minimal length of time required to complete the project Increasing the time required for any critical activity will also increase the time necessary to complete the project When drawing weighted digraphs for problems involving critical paths the following conventions are used:  The edges (or arcs) represent the activities  The vertices (nodes) represent events – the start/finish of one or more activities is called an event  An edge should lead from a start vertex to represent each activity that has no predecessor  A vertex (called the finish node) representing the completion of the project should be included in the network  Two nodes can be connected directly by, at most, one edge Sometimes a dummy activity is introduced to a digraph – a dummy activity is an artificial activity of zero time duration added to a project diagram to ensure that all predecessor activities are properly accounted for The earliest starting time (EST) is the earliest time an activity can be commenced The activities without predecessors have earliest start time 0. You work from left to right through the digraph to determine the earliest starting time for an activity The latest starting time (LST) is the latest time an activity can be left if the whole project is to be completed on time Latest starting times are established by working backwards through the network Start in the boxes to the far right with both entries equal and then subtract duration times to move back through the diagram – to obtain the other final event times, subtract the activity time (working backwards) If the difference between the earliest starting time for an activity and the latest starting time for an activity is 0, the activity is critical The float or slack of a non critical activity is the amount by which the latest starting time is greater than its earlier starting time: It may be possible to speed up a project by completing some activities more quickly – this is called project crashing When writing the critical path for the graph below, it is to include the dummy activity, written as B -C–E–G–T Whilst when writing the floats for the non critical activities, exclude the dummy activities ( ) When determining the predecessors for an activity, if the predecessor is a dummy activity, trace it to the predecessor of the dummy activity (the predecessor of activity C is A and B, whilst the predecessor of activity E is C and D)

The Hungarian algorithm is an algorithm used to solve allocation problems (eg. Determining minimum cost) When the cost in dollars of an action is recorded in a table, it is called a cost matrix There are four steps to applying the Hungarian algorithm:

Linear: Chapter 16 – 18
Through any two points it is possible to draw only a single straight line – therefore a straight line is defined by any two points on the line A gradient is defined as the slope of a line – the symbol used for gradient is m or [where (x2, y2) and (x1, y1) are two points on the line] The gradient of a line that slopes upwards from left to right is positive The gradient of a line that slopes downwards from left to right is negative The gradient of a horizontal line is zero, since =0 The gradient of a vertical line is undefined, since 2 =0 The general equation of a straight line is y = mx + c, where m is the gradient of the line This form, expressing the relation in terms of y, is called the gradient form The y-axis intercept is equal to c [but alternatively, is found by letting x=0] The x-axis intercept is found by letting y = 0, and solving for x If the equation for the straight line is not written in gradient form, the graph may be transposed to aid sketching If the value of m is the same for two rules, then the lines are parallel – for example, consider the lines with the rules y = 2x + 3 and y = 2x – 4, they are parallel because they both have the same gradient, m = 2. The equation of a straight line can be found if the gradient and the y-axis intercept are known If we are missing the y-intercept, gradient or x-intercept, we can still find the equation to the line Using our knowledge of , we can transpose it to establish: The equation of a straight line can also be found from the graph by reading off two points and using the formula If m = 0, the line is horizontal and the equation is simply y = c, where c is the y-axis intercept If the line is vertical, the gradient is undefined and thee rule is given as x = a, where a is the x-axis intercept When the y-axis intercept = 0, the equation is y = mx Often we encounter a linear relation that is not expressed in the form y = mx + c, but rather, in an alternative standard notation, ax + by = c, which is sometimes referred to as intercept form Although gradient form is required to find the gradient, it is often convenient to work in intercept form A convenient way to sketch graphs of straight lines is to plot the two axes intercepts A linear equation that contains two unknowns does not have a single solution If the graphs of two equations are drawn on the same set of axes, and they are non-parallel, the lines will intersect at one point only – hence there is one pair of numbers that will satisfy both equations simultaneously For each price of a product a corresponding quantity of the product will be demanded (purchased) during some time interval – the quantities of a product that can be demanded or supplied can be represented graphically The point where two lines intersect is called the breakeven point A graph made up of two or more parts of the different straight line graphs, each representing different intervals on the x-axis, is called a line segment graph (segmented linear graphs) Step graphs are line segment graphs were each segment is horizontal For step graphs, open dots means , whilst coloured dots mean Consequently, when reading a value off a step graph, always read from the coloured dot, not the open one Finding a value from a graph that is within the range of values used to construct the graph is called interpolation Finding a value from a graph that is outside the range of values used to construct the graph is called extrapolation

these are hyperbolas (eg. these are parabolas with the vertex as the origin (eg.For the graph of :  When n = 1. When . ) – the graph approaches both the x-axis and the y-axis but does not cross either lines – these lines are referred to as asymptotes  When n = 2. these are straight lines through the origin  When n = -1.

x 2) the region that complies with this inequality (in this case. a linear graph can be obtained by graphing . where m becomes 3 Linear programming is often used to determine a solution to a problem to maximize profit or minimize cost – these problems are called optimization problems When the equation to a line is an inequality (eg. For parabola o the possible y values are all positive real numbers o it is symmetrical about the y-axis – the line about which the graph is symmetrical is called the axis of symmetry which is a vertical line which is located through the turning point (middle) o the graph has a vertex. or turning point. at the origin (0. these are truncus.0) o the minimum value of y is 0 and it occurs at the turning point When n = -2. a family of basic cubics is formed Consider the line for . all the values of x greater than two) is shaded and consequently referred to as the required region . named for their shape that resembles a trunk (elephant/tree)  When n = 3.

it is easiest to use (0. if two corners give the same maximum or minimum value. and is included (continual line) if are used When testing to find the required region.0) if it is not part of the boundary Regions can be defined by multiple inequalities – where the required region is where the required regions of the two lines overlap The required (shaded) region is also called the feasible region Points on the boundary (if continual line – should always be) are included in the feasible region An objective function is used to describe a quantity that you are trying to make as large as possible or as small as possible subject to some constraints [the objective function is always given if needed] In linear programming problems. the aim is to find the maximum and minimum value of an objective function for a given feasible region – to help us do this. we can make use of the corner point principle In linear programming the maximum and minimum value of a linear objective function will occur at one of the corners of the feasible region – note.The convention is that if the boundary is included it is represented with a continuous line ( ) and if it is not included is represented with a dotted line ( ) The boundary is not included (dotted line) if are used. all points along a line joining the two points will also give the same maximum and minimum .

whether a person smokes WARNING: It is not the variable name itself that determines whether the data is numerical or catergorical. we must determine the variable type Some values represent qualities or attributes (sex: female/male) Variables that represent qualities are called categorical variables Others variables represent quantities (height: in centimeters. 3 = overweight. the order in which the categories Frequency (Female/Male) are listed is quite arbitrary.5 had been first. it would make since to list them in Male 5 45. salary (high-low). number of children in a family.3 etc. number of people who live in your city/area Examples of continuous data: Daily temperature. the categories Female 6 54. city of residence. 2 = normal weight. age (years).Statistics: Chapter 1 .) – counting leads to discrete data values because it results in values such as 0. there is no natural Sex Count Percentage order (it doesn’t matter). – there can be nothing in between (only whole numbers) – typically discrete numerical variables arise when asked how many? Continuous numerical variables represent quantities measured rather than counted (as even though we might record a person’s height as 179 cm. time spent watching TV.0  The total count should always equal the total number of observations (11) – however.9 or 100. number of cars in a supermarket car park. however.5 and 179.) Once the variables are determined. the number of cigarettes smoked per day. second and third.1 is sometimes obtained – this is due to the rounding error – totaling the count and percentages helps check on your counting and percentaging . weight recorded as 1 = underweight.2. length of bananas. their units of measurement and the values they take (eg. if percentages are rounded to one decimal place a total of 99. really. along with how frequently each occurs is called a frequency distribution Frequency tables are primarily used for categorical data – non numerical data There are two things to remember in constructing a frequency table:  In setting up the table. it could be any values between 178. for example. weight: in kilograms) Variables that represent quantities are called numerical variables Numerical variables come in two types: discrete and continuous Discrete numerical variables represent quantities that are counted (the number of mobile phones in a house etc. it is the way the data variable is recorded (weight recorded in kilograms is a numerical variable. it is difficult to identify any patters or trends in the raw data – we first need to organize the data into a more manageable form – a statistical tall we use for this purpose is a frequency table A frequency table is a listing of the values a variable takes in a data set along with how frequently each occurs Frequency can be recorded as a :  Count: the number of times a value occurs  Percent: the percentage of times a value occurs ( ) A listing of the values a variable takes. salary ($) Examples of categorical data: Eye colour.1. sex (M or F).7 Statistics is a science concerned with understanding the world through data – the first step in this process is to put the data into a form that makes it easier to see patterns and/or trends In a data et we call the things about which we record information variables An important first step in analysing any data set is to identify the variables involved.5 that order Total 11 100. the TV channel most watched. weight (kg).4cm – we have rounded off for convenience or to match the accuracy of the measuring device) [decimal numbers] Examples of discrete variables: Shoe size. height (cm). if. pulse rate (beats/minute) etc. is a categorical variable) With a large number of data values.

the lengths of each of the segments in the bar are determined by the percentages – when this is done.5-9. 60. as you don’t know which interval to put 5. whilst continuous data differs Some variables can only take on a limited range of values (number of children etc.9% were found to have a moderate climate.) – which is why it makes sense to list each value individually when forming a frequency distribution/table. usually no more than four or five. this should be stated  The order of occurrence of each category and its relative importance In commenting on these features. 5-10.The process of forming a frequency table for a categorical variable:  Displays the data in a compact form  Tells us something about the way the data values are distributed (the pattern of the data) Once categorical data has been organized into a frequency table. is useful for gaining an overall view of a frequency distribution of a categorical variable. the variable can take a larger range of values (0-100) – listing all possibilities would be tedious and produce a large and unwieldy display – to solve this problem.percentage segmented bar charts are most useful when we come to analyse the relationship between two categorical variables Although it is more common for frequency tables to be used for categorical data. it is also advisable to indicate at the beginning total number of cases involved – example: “The climate type of 23 countries was classified as being cold. the height of the bar is always 100 . In describing a bar chart we focus on two things:  The presence of a dominant category (or group of categories) in the distribution – this is given by the mode – if there is no dominant category. moderate or hot. In other cases. The majority of the countries.in combination with a frequency table.” A variation on the standard bar chart is the segmented or stacked bar chart In a segmented bar chart. because it becomes difficult to distinguish between components in a percentage segmented bar chart.15-19.0% were found to have a cold climate. it can be used for numerical data also – for discrete data the process exactly mirrors that for categorical data.1% were found to have a hot climate. 10 or 15 in In this case. aft.99.00 and 8. it is usual to support your conclusions with percentages – when quoting percentages.10-14.99 would be used . it is common practice to display the information graphically to help identify any features that stand out in the data – we often use a bar chart The bar chart pictures the key information in a frequency table – it is used for categorical data – In a bar chart:  Frequency (or percentage frequency) is shown on the vertical axis  The variable being displayed is plotted on the horizontal axis  The height of the bar (column) gives the frequency (or percentage frequency)  The bars are drawn with gaps between them to indicate each value is a separate category  There is one bar for each category One of the features of a data set that is quickly revealed with a bar chart is the mode or modal category This is the most frequently occurring value or category – this is given by the category with the tallest bar Mode is especially important in answering questions like which is the most frequently watched TV station between 6. Etc. while 13.00 pm? or what are the times when a supermarket is in peak demand? (morn. the bars are stacked on one another to give a single bar with several components The lengths of the segments are determined by the frequencies – when this is done the height of the bar gives the total frequency – segmented bar charts should only be used when there are a relatively small number of components.99. between five and fifteen intervals are used Note that the intervals are defined so that it is quite clear into which interval each data value falls – for example: you cannot define intervals as 1-5. the intervals of 1-4. 15-20…. we group the data into a smaller number of convenient intervals – There are no hard and fast rules for the number inters but normally. Of the remaining countries.99. night) A bar chart . the so called big picture. 26. 10-15.

9 6 26.9 The process of forming a frequency table for a numerical variable:  Orders the data  Displays the data in a compact form  Tells us something about the way the data values are distributed (the pattern of the data)  Helps us identify the mode (most frequently occurring value or interval of values) The frequency histogram. or histogram for short.0 Total 23 99.9 3 13.9499  The modal interval is 40.0 because it contains all values from 34. the intervals start and end halfway between values – empty classes or missing discrete values have bars of zero height  The height of the bar gives the frequency (usually the count. eight (34.9 1 4.1 40.9500….9%) of the countries have working hours that fall into this interval Frequency Average hours worked Count Percentage 30.There are two things to note from this continuous variable frequency table:  The intervals in this example are of width five – for example.0-39.0-34. where there are gaps between values.9 8 34.9 5 21. to 39.9 is an interval of width 5. is a graphical way of presenting the information in a frequency table for numerical data – in a frequency histogram:  Frequency (count or percent) is shown on the vertical axis  The values of the variable being displayed are plotted on the horizontal axis  For continuous data.7 50.9 hours.8 45.0 .3 35.0-39. the interval 35.44. but it can equally well be the percentage) Construct a histogram for continuous data: Construct a histogram for discrete data: .0-44. each bar in a histogram corresponds to a data interval – for discrete data.0-49.0-54.

do some data values tend to occur much more frequently than others. they may indicate data values that are very different from the rest of the values .It is relatively quick to construct a histogram from a prepared frequency table – however. the distribution is said to be symmetric A single peaked symmetric distribution is characteristic of the data that derives from measuring variables such as people’s heights. if we were studying the distance the discus is thrown by Olympic discus throwers. intelligence test scores or any data for which the values vary evenly around a central value [the histogram for average hours worked would be considered approximately symmetric] The double peaked distribution is symmetric about the dip between the two peaks . for example. showing that all values in the distribution occur with approximately the same frequency? If a histogram is single peaked. does the histogram tail off evenly on either side of the peak? If so. a person’s age recorded as 356 Alternatively. if you only have raw data it is a very slow process because you have to construct a the frequency table first In describing a histogram we focus on three things:  Shape and outliers (values of data that appear to stand out from the rest)  Centre  Spread How is the data distributed? Is the histogram peaked. we would expect a bimodal distribution if both male and female throwers were included in the study) Sometimes a histogram tails off primarily in one direction – such distributions are said to be skewed If a histogram tails off to the right we say that it is positively skewed The distribution of workers in large organizations tends to be positively skewed If a histogram tails off to the left we sat that it is negatively skewed Outliers are any data values that stand out from the main body These are data values that are atypically high or low Outliers can indicate errors made collecting or processing data. that is. or is it relatively flat.A histogram which has two distinct peaks indicates a bimodal (two modes) distribution A bimodal distribution often indicates the data has come from two different populations (for example.

Center of a distribution refers to the middle of the distribution of the data In this histogram (left) we can see that they are identical in shape.But what do we mean by the spread of a distribution? For a histogram we will take it to be the maximum range of the distribution It should be noted that with grouped data. outliers. is it narrow? This would indicate that most of the data values in the distribution are tightly clustered in a small region. Or is the peak broad? This would indicate that the data values are more widely spread apart . is an alternative to the histogram – it is particularly useful for displaying small to medium sized sets of data (up to about 50 data values) and has the advantage of retaining all the original data values – this makes it useful for further computations A stem plot is also a very quick and easy way to order and display a set of data by hand Like a histogram. The spread of the distribution is 1020 phones/1000 people. the scores tend to bunch up at one end – however. the distribution of the number of phones per 1000 people is positively skewed. but differ in where they are located along the x axis In statistical terms. centre and spread of the distribution . it is acceptable to state that the centre of a distribution lies within the interval of x – y Example analysis of a histogram: For the 85 countries. it is difficult to precisely determine the location of the centre of distribution from a histogram – so when working with grouped data. we say that the distributions are centered at different points along the axis The middle of a symmetric distribution is reasonably easy to locate by eye – but for skewed distributions it is more difficult to estimate the middle of a distribution by eye The middle is not halfway between the extremes because. or stem plot for short. The centre of distribution lies somewhere in the interval 170-340 phones/1000 people. A stem and leaf plot. the midpoint lies on the line that divides the histogram into two equal areas If the histogram is single peaked. I a skewed distribution. if we imagine a cardboard cutout of the histogram. stem plots give information about the shape. There are no outliers.

the media will be an actual data value – for an even number of data values.14. the third 1 represents the interval 14 . the value which has an equal number of data values either side – for an odd number of data values.19 In a stem plot with its stem split into halves. enabling more precise estimates For the centre of a stem plot. the middle value is the value – thus.13. the fourth 1 represents the interval of 16 . each data value is separated into two parts: the leading digit(s) form the stem. The centre of the distribution is 30 marks and the distribution has a spread of 45 marks. In some instances.19 . enter the stems to the left of a vertical dividing line. n.One of the stem plots advantages over a histogram in describing distributions is being able to see all the actual data values – this enables the centre and the range of the distribution to be located more precisely. and the trailing digit becomes the leaf – for example. while the bottom 1 represents the interval of 15 – 19 In a stem plot with its stem split into fifths. using the simple process outlined above produces a stem plot that is too bunched up to give us a food overall picture of the variation in the data – this is often the case when the data values all have the same first digit or the same one or two first digits – for example (VCE results out of 20): 2 12 13 9 18 17 7 16 12 10 16 14 11 15 16 15 17 Using the above process to form a stem plot we end up with a bunched up stem plot: 0|2 7 9 1|0 1 2 2 3 4 5 5 6 6 6 7 7 8 When this happens. it also enable the clear identification of any outliers.17 while the bottom 1 represents the interval of 18 . and the leaves for each data point to the right – usually we first construct an unordered stem plot by systematically plotting each data point as listed in the set data – from the unordered stem and lead plot an ordered stem plot is them easily obtained In an ordered stem plot the leaves increase in value as they move away from the stem – it is usually an ordered stem plot that we want because an ordered stem plot makes it easy to find the key values Stem plots are just like histograms except that you can see all the data values. the second 1 represents the interval 12 . in a stem and leaf plot. the data values represent 25 and so on To construct a stem plot. the 1 represents the interval 10 . the distribution of marks is approximately symmetric with an outlier. In a stem and leaf plot. the middle value is the vale – thus the median will lie between the two data values For the spread (range) subtract the smallest data value from the largest data value Data values that stand out from the main body are called outliers – their values can be read directly from the plot Example of an a description of the distribution of a stem plot: For the 23 students. the top 1 represents the interval of 10 . n. the top 1 represents the interval 10 . the stem plot scale can be stretched out by splitting the stem – generally the stem is split into halves or fifths – in this case it would occur as follows:    In a stem plot with a single stem.11. The outlier is a mark of 60. count up from either end of the distribution until you find the middle value.15.

indicating that the mark of 2 is an outlier: A back to back plot has a single stem with two sets of leaves as shown – the real value of back to back plots is that they are a useful tool for comparing the distribution of two sets of data values for the same variable The simplest way to display numerical data is in a dot plot a dot plot consists of a number line with each data point marked by a dot – when several data points have the same value. we can see that the distribution is centered at 22 years with a spread of 11 years Type of data Categorical Numerical Graph Bar chart Segmented bar chart Histogram Stem plot Dot plot Qualifications on use Not too many categories (4 – 5 maximum Best for medium to large data sets ( ) Best for small to medium sized sets ( ) Only suitable for small data sets ( ) Just a few numbers obtained from the data can be used to describe the essential features of distribution We call these numbers summary statistics The most commonly used types of summary statistics may be classified as:  Measures of centre (about what point is the distribution centered)  Measures of spread (how are the scores in the distribution spread out) . usually there is little we can say about the shape of the distribution of the dot plot because there are not sufficient data points for an patter From the dot point above. the points are stacked on top of each other Like stem plots. dot plots are a great way for displaying small data sets and have the advantage of being very quick to construct by hand – they are best when the data is relatively close together Dot plots are interpreted in much the same way as stem plots – however.Using a split stem plot to display the test marks can show features not revealed by a standard plot – this can seen in the next plot with stem split into fifths.

the median. the median divides the data set up into for equal groups The IQR is a measure of spread of a distribution that includes the middle 50% of observations – since the upper 25% and the lower 25% of observations are discarded. irrespective of the way in which the data values are spread out in between – however. and the smallest and largest values of a distribution. write the data values down in order and mark in the median and quartiles – if correct. the range is useful to know as it gives us an indication of the absolute spread of the distribution Just as the median is the point that divides a distribution in halp. M.The median is the midpoint of a distribution: 50% of values in the data set are less than or equal to the media The median is the middle value in a data set – its value is found by listing all the data values in numerical order We then find the value that divides the distribution into two equal parts – for small data sets the median can easily be located by eye but for larger data sets the following rule for locating the median is helpful: For n ordered data values. written in the order: is known as a five number summary The five number summary can be used to construct a new graph known as the box plot . so that: range = largest data value – smallest data value The range as a measure of spread has limitations – because the range depends only on the two extreme values in a set of data it is not always an informative measure of spread – for example. any two sets of data with the same highest and lowest values will have the same range. the interquartile range is generally no affected by the presence of outliers – this makes it a more useful measure of spread than range A listing of the media M. R. the largest and smallest values in a data set might be outliers and not at all typical to the rest of the values – furthermore. the quartiles Q1 and Q3. is the simplest measure of spread of a distribution – it is the difference between the larfest and smallest values in the data set. is located at the position The range. quartiles are the points that divide a distribution into quarters – the symbols of are used to represent the quartiles (note ) The interquartile range (IQR) is defined to be the spread of the middle 50% of data values so that: so that To calculate the IQR it is necessary to first calculate the quartiles Q1 and Q3 – which is as straight forward as:  Q1 is the midpoint of the lower half of the data values  Q3 is the midpoint of the upper half of the data values To check that the quartile you obtain are correct.

5 IQR’s from the ends of the box A symmetric distribution is evenly spaced around the median – there is also a strong tendency for data values to cluster around the centre of the distribution rather than at the extremes .The box plot is an extremely powerful tool for describing the distribution of data In its simplest form.5 IQR’s away from the end of the box are classified as possible outliers (possible in that it may be that they are just part of a distribution with a very long tail and we do not have enough data to pick up other values in the tail)  The whiskers end at the highest and lowest data values that lie within 1. spread and general shape of distribution Box plots can be drawn either vertically or horizontally (predominantly horizontally) The box plot with outliers is a more sophisticated form of the box plot and is designed to identify any outliers that may be present in the data: Two new things to note in a box plot with outliers are that:  Any points more than 1. the box plot (or box and whisker plot as it is sometimes known ) is a graphical version of a five number summary – it is a very compact way of displaying the location.

we can see that shape of the – we can also see whether or not there are any outliers – furthermore.Positive skewed distributions are characterized by a cluster of data values at the left hand end of the distribution with a gradual tailing off to the right Negatively skewed distributions are characterized by a cluster of data values at the right hand side of the distribution with a gradual tailing off to the left Less interesting but of practical importance. this is given by the length of the box plus the whiskers – if there are outliers. the IQR – the second measure of spread given by a box plot is the range – when there are no outliers. the spread of the distribution can be seen in two ways – the first given by the length of the box. the centre of the distribution is clearly identified and given a value by the median – finally. these are also included in determining the value of the range . an outlier might signal an error in the data Because of the wealth of information contained in a box plot. it is an extremely powerful tool for describing the distribution – at a glance. which corresponds to the spread of the middle 50% of values.

Σ. 8 The median remains at M = 3. consider the set: 2. the mean is the balance point of the distribution For approximately symmetric distributions. 3. and vise versa Outliers drag the mean away from the body of data . . that is. the value of mean will be relatively larger. we will introduce a shorthand notation that enables complex statistical formulas to be written out in a compact form – in this notation we use:  The Greek capital latter sigma. however. there are two data points to the left of the mean. (pronounce x bar) to represent the mean of the data values  An n to represent the total number of data values The rule for calculating the mean then becomes Whereas the median lies at the midpoint of a distribution. later you will meet other rules for calculating statistical quantities that are extremely complicated and hard to write out in words To overcome this problem. and two to the right This is a generally characteristic of symmetric distributions However consider the data set: 2. 3. the mean is the balance point of a distribution Note that in this case. the distribution is asymmetrical.The mean of a set of data is what most people refer to as the average . the balance point of distribution is also the point that splits the distribution in half. 4 The mean of the set of data is given by: From a pictorial point of view. 3. the mean and median values will/may vary quite significantly If a graph is negatively skewed. both the median and mean will be approximately equal in value If. the mean and the median coincide. as a shorthand way of writing sum of  A lower case x to represent a data value  A lower case x with a bar. but: Note that the mean is affected by changing the largest data value by that the media is not Because the rule for the mean is relatively simple. it is easy to write in words – however. 3.the mean of a set of data is given by: For example.

it is said to be a resistant statistic – for this reason. is: where n is the number of data values (sample size) and is the mean . the median is frequently used as a measure of centre when the distribution is known to be cleared skewed and/or likely to contain outliers Likewise. if a distribution is symmetric. s.Because the value of the median is relatively unaffected by the presence of extreme values in a distribution. particularly in the area of statistical inference To measure the spread of a data distribution around the median (M) we use the interquartile range (IQR) To measure the spread of a data distribution about the mean ( ) we use the standard deviation (s) The formula for the standard deviation. there will be little difference in the value of the mean and median and we can use either – in such circumstances. the mean is often preferred because:  It is more familiar to most people  More can be done with it theoretically. the quartiles Q1 and Q3 are resistant statics and are used as indicators of spreads in situations However.

Let us say that all the data values lie within two standard deviation of the mean (as we know 95% do) – if this is true then to estimate standard deviation: Many data sets that arise in practice are roughly symmetrical and have an approximate bell shape as shown: Data distributions that are approximately bell shaped can be modeled by a normal distribution In normal distributions.7% of the observations lie within three standard deviation of the mean Around 68% of the data values will lie within one standard deviation (SD) of the mean: This also means that 32% of values lie outside the region – as the distribution is approximately symmetrically we can also say that around 16% of values lie in each of the tails (the blue section) . two or three standard deviations of the mean – this gives rise to what is known as the 68 – 95 – 99. the peerage of observations that lie within a certain number of standard deviations of the mean can always be determined In particular we are interested in the percentage of observations that lie within one.7% rule For a normal distribution. approximately:  68% of the observations lie within one standard deviation of the mean  95% of the observations lie within two standard deviation of the mean  99.

Around 95% of the data values will lie within two standard deviations (SD) of the mean: This also means that 5% of values lie outside the region – as the distribution is approximately symmetrically we can also say that around 2.7% of the data values will lie within three standard deviations (SD) of the mean: This also means that 0.3% of values lie outside the region – as the distribution is approximately symmetrically we can also say that around 0.15% of values lie in each of the tails (the blue section) In addition.5% of values lie in each of the tails (the blue section) Around 99. the man and the median coincide so that: 50% of the data values will lie above the mean and 50% of values will lie below the mean: . because the normal distribution is symmetric.

subtract the mean from the data value and then divide the result by the standard deviation – That is: Standard scores can be both positive and negative:  a positive z score indicates the data value it represents lies above the mean  a zero standardized score indicates the data value is equal to the mean  a negative z score indicates the data value it represents lies below the mean Consequently:  a z score of 1 means the value is one standard deviation above the mean  a z score of 2 means the value is two standard deviations above the mean  a z score of 3 means the value is three standard deviations above the mean  a z score of -1 means the value is one standard deviation below the mean  a z score of -2 means the value is two standard deviations below the mean  a z score of -3 means the value is three standard deviations below the mean In statistics. the larger group of people/things that you would like to extend you conclusions to is a population The group of subjects that you choose to work with is your study is called the sample The easiest way of obtaining a representative sample from a population is to select a simple random sample (SRS) A population parameter is a number describing a population A sample statistic is a number that can be calculated from sample data We can combine two frequency tables to determine whether there is any noticeable relationship (between sex and attitude towards gun control) – to do this we need to create a two way frequency table – the process of forming a two way frequency table is called crosstabulation .Because of the additional insight provided by being able to relate standard deviations to percentages. it is common to transform normally distributed data into a new set of units which show the number of standard deviations each data value lies from the mean of the distribution This is called standardizing and these data values are called standard or z scores To obtain a standard score for a data value.

2%. we call the variable that depends on the other the dependent variable (DV) and the variable it depends on the independent variable (IV) In two way frequency tables.9% to 85. but not the other way around – in such situations. the regions shaded in blue (above) are called the margins of the table  the numbers in the right margin are called the row sums  the numbers in the bottom margin are called the column sum  the number in the right hand corner is called the grand sum (if constructed correctly grand sum = 100.2% of males in the sample were for gun control compared to 71.4% to 55. This indicates that a person’s attitude to gun control is related to their sex Example analysis report for unrelated variables: From the percentage table we can see that the percentage of males and females in support of banning mobile phones in cinemas was similar. it is conventional to let the categories of the dependent variable define the rows of the table and the categories of the independent variable define the columns of the table In a two way frequency table. we would expect roughly equal percentages for both sexes Example analysis report for related variables: From the percentage table. to answer our question we are to rely on column percentages Column percentages are determined by dividing each of the cell frequencies by the column totals Percentaging a (this) table enables us to compare variables equally (attitudes of M/F on equal footing) From the table we can see that 55. 87.8%. we see that a higher percentage of females were for gun control than males. or are men really more in favour of gun control than women? – to help us answer this question we turn our table entries into percentages Although there are multiple ways to construct a percentage two way frequency table.When studying the relationships between variables.0) The regions in the table shaded in purple are called the cells of the table It is the numbers in these cells that we look at when investigating the relationship between two variables From the example above we can conclude that:  32 males are for gun control  30 females are for gun control This information tells us that more men are in favour of gun control than women (but is due to the fact that there were more men in the sample. not %’s This percentage table indicates that a person’s attitude to gun control depends on their sex – we can say that these two variables (sex and attitude to gun control) are related or associated If attitude to gun control and sex were unrelated. it is sometimes clear that one of the variables might depend on the other. This indicates that a person’s support for banning mobile phones was not related to their sex . 71.4% of women This means that the females in the sample were more supportive of gun control than the males This reverses what the frequency table told us – it is easy to be misled if you just compare frequencies.

when dealing with variables with more than two categories. however. attitude to gun control (GC) is related to sex If there was no relationship we would expect the bottom segments in each bar to be roughly equal in length It is easy to use a two-by-two table to see whether variables are related by comparing percentages. it is not always so easy to identify trends – in such circumstances.This process can be made easier by using a percentaged segmented bar chart to display percentages graphically Each column in the bar chart corresponds to a column in the purple shaded region of the table Each segment corresponds to a cell in the table From the segmented bar chart. the segmented bar chart is useful a useful aid . we can see clearly that a greater percentage of females than males favour GC This indicates that for this group of people.

This indicates that smoking is related to level of education The statistical tool that we use to investigate the relationship between a numerical variable and a categorical variable is a series of parallel box plots – in this display there is a box plot for each category of the categorical variable – relationships can then be identified by comparing the distribution of the numerical variable in terms of shape. the use of all three gives a more complete description of this relationship The first step in investigating the relationship between two numerical variables is to construct a scatterplot In a scatterplot.From the segmented bar chart. For example. it is clearly seen that as education level increases. the numerical variable is the DV and the categorical variable is the IV) Comparing medians: From the parallel box plots we can see that median salaries increase with age group .4% for university. Note – any one of these reports by themselves can be used to claim that there is a relationship between two variables. however. looking at the bottom segment in each column. each point represents a single case (eg. there is a decrease in the percentage of past smokers – thus we can conclude that there is a relationship between smoking and education level in this sample – an example report/analysis is as follows: From the percentage frequency table we can see that the percentage of smokers clearly decreases with education level from 33.9% for year 9 or below. This is an indication that the shape of the distribution of salaries is related to age group. This is an indication that typical salaries are related to age group Comparing IQRs and/or ranges: From the parallel box plots we can see that spread of salaries increased with age. a country) – the horizontal or x coordinate of the point represents the independent variable. more outliers begin to appear. the IQR increased from around $12000 for 20-29 year olds to around 20000 for 50-65 year olds. We can also see that with increasing age. It is approximately symmetrical for the 20-29 year olds and becomes progressively more positively skewed with increasing age. centre and spread (for example parallel box plots that show the salary distribution of four different age groups – note that in this situation. to 18. indicating salaries well above normal. whilst the vertical or y coordinate represents the dependent variable The scatterplot is completed by plotting all of the points from the table of values onto a graph . This is an indication that the spread of salaries is related to age Comparing shapes: From the parallel box plots we can see that the shape if the distribution of salaries changes with age. from around $24000 for 20-29 year olds to around $32000 for 50-65 year olds.

When constructing a scatterplot it is conventional to use the vertical or y axis for the dependent variable (DV) and the horizontal or x axis for the independent variable (IV) When interpreting scatterplots. we must first look to see if there is a clear pattern in the scatterplot Having found a clear pattern. there are several things we look for in this pattern of points These are: direction and outliers (if any). form & strength .

We are also looking to see whether the pattern in the points has a linear form If the points in a scatterplot can be thought of as random fluctuations around a straight line. then we say that the scatterplot has a linear form – if the scatterplot has a linear form then we say that the variables involved are linearly related .

moderate or strong If you are using the value of the correlation coefficient as a measure of the strength of a relationship. is used If there is a less than perfect linear relationship. we say that there is a perfect linear relationship – if there is no fit at all we say that there is no relationship to measure the strength of a linear relationship the correlation coefficient. but it is clearly non-linear – the scatterplot shows that while level of performance on a task will increase with practice. there comes a time when the performance level will not improve substantially with extra practice – while non linear relationships exist. then the correlation coefficient r has a value between -1 and +1 or Our reason for estimating the value of the correlation coefficient is to give a measure of the strength of the linear relationship – when doing this. r. the scatterplot opposite clearly has a non linear form – this is a plot of performance level against time spent on practicing a task – there is a relationship between performance level and time spent in practice. we sometimes find it useful to classify the strength of the linear relationships as weak.the correlation coefficient can five a misleading indication as to the strength of the linear relationship if there are outliers present . then you are implicitly assuming  The variables are numeric  The relationship is linear  There are no outliers . many of the relationships we meet in practice are linear or maybe made linear by transforming the data The strength of a linear relationship is an indication of how closely the points in the scatterplot fit a straight line – if the points in the scatterplot lie exactly on a straight line.By contrast.

5 and 0.24 Weak negative relationship r between 0. then r is given by: In this formula.49 Moderate negative relationship r between 0.75 and 0. and we have n observations. gives a numerical measure of the degree to which the points in the scatterplot tend to cluster around a straight line – formally we call the two variables we are working with x and y. and sx are the mean and standard deviation for the x values and and standard deviation of the y values and sy are the mean .5 and 0.99 Pearson’s correlation coefficient.99 Moderate positive relationship r between 0.25 and 0.49 No relationship r between 0.24 and 0. r.75 and 0.25and 0.74 Weak positive relationship r between 0.74 Strong negative relationship r between 0.Strong positive relationship r between 0.

lifestyle.64) of the variation is peoples weights can be explained this way – the rest of the variation (36%) will be explained by other factors such as sex. and r2 = 0.8. shorter people tend to be lighter – the coefficient of determination tells us that 64% (as r = 0.If two variables are related. we find that both their weights and their heights will vary – one explanation for the variation in peoples weights is that their heights vary – taller people tend to be heavier. it is possible to estimate the value of one variable from the value of the other The degree to which one variable can be predicted from another linearly related variable is given by a statistic called the coefficient of determination – the coefficient of determination is calculated by squaring the correlation coefficient: In technical terms the coefficient of determination tells us that r 2 x 100 percent of the variation in the dependent variable (DV) is explained by the variation in the independent variable (IV) If we take a group of people. age and build While you might establish a relationship between two variables. this in itself is not sufficient to imply that a change in one of the variables will cause a change in the other 0 many other explanations can be found that that might explain both variables – thus two variables may vary together without directly being the cause of the other and we must be aware of not reading too much into any relationship The process of fitting a straight line to bivariate data is known as linear regression – the aim of linear regression is to model the relationship between two numerical variables by using a simple equation of a straight line To do this we rely on two methods:  The least squares method  The three median method The most common approach to fitting a straight line to data is to use the least squares method – this method assumes that the variables are linearly related and works best when there are no clear outliers in the data . for example.

it minimizes the sum ( ) We minimize the sum of the squares of the residuals and not the sum of the residuals as the sum of the residuals is always zero for the least squares line – the least squares line is like the mean – it balances out data values on either side of itself – some residuals are positive. that is. some negative. where: WARNING: to calculate the least squares line it is vital that you correctly determine the DV and the IV .The least squares line is the line where the sum of the squares of the residuals is as small as possible. and in the end they add to zero The least squares line is the line that minimizes the sum of the squares of the residuals the equation for the least squares regression line is given by  the slope (b) is given by:  Here:    the intercept (a) is then given by r is the correlation coefficient sx and sy are the standard deviations of x and y are the mean values of x and y .

A full regression analysis involves several processes. which include:  Constructing a scatterplot to investigate the nature of the relationship between the variables  Calculating the correlation coefficient to give measure of the strength of the relationship  Determining the equation of the regression line  Interpreting the coefficients of the least squares line  Using the regression line to make predictions  Using the coefficient of determination to give a measure to the predictive power of the linear relationship  Using a residual plot to test the assumption of linearity  Writing a report on your findings The scatterplot and correlation coefficient: The least squares regression line: .

For the regression equation y = a + bx:  The slope. then y decreases as x increases  They intercept predicts the value of y when x = 0 We can also use the regression equation as a means of prediction . b. predicts the change in y when x changes by one unit: o If the slope b is positive. then y increases as x increases o If the slope b is negative.

The clear curved pattern in the residual plot show below confirms this assertion. r = -0. The equation of the least squares regression line is: life expectancy = 105. an average life expectancy decrease by 1. The slope of the regression line predicts that. the relationship between the yield of potato plot and its length is clearly non linear – see the scatterplot below. the horizontal zero line helps us to orient ourselves This line corresponds to regression in the previous scatterplot From the residual plot. Essentially they are randomly scattered around the zero regression line. we see that there is no clear pattern in the residuals. Thus from the residual plot we can conclude that: The assumption that there is a linear relationship between life expectancy and birth rate is confirmed by the residual plot In contrast to the relationship between life expectancy and birth weight. linear relationship between life expectancy and birth rate.1% of the variation in life expectancy is explained by the variation in birth rate. The coefficient of determination indicates that 65. the predicted value – there is also a random part that cannot be predicted by the regression line – this random part is just the residual value.44 x birth rate.807. The final step in regression analysis is to report on your findings The report is in a form that is suitable for including a statistical project – note that an interpretation of the intercept of the regression equation is not included for this example as it has no meaningful interpretation From the scatterplot we see that there is a strong negative. thus we can write that: or If we continue to calculate the residuals. A residual plot shows no clear pattern and confirms that the use of a linear equation to describe the relationship between life expectancy and birth rate is appropriate . All that is left after fitting the regression line is random variation around the line.A key assumption made when calculating a least squares regression line is that the relationship between the variables is linear – one way of testing this assumption is to plot the regression line on the scatterplot and see how well a straight line fits the data – however. This confirms our original assumption that a linear relationship between life expectancy and birth weight is reasonable. There are no obvious outliers. we will find that some are positive and some are negative What we hope is that there is no clear pattern to the residuals – to see this we constrict a residual plot A residual plot is a plot of the residual value for each data vale against the independent variable – because the mean of the residuals line is always zero.4 – 1. a better way is to use a residual plot as this plot will show even very small departures from linearity When we fit a least squares line we assume that part of each data value can be predicted by the regression line.44 years for an increase in birth rate of 1 birth per 1000 people.

it is less sensitive to outliers than the least squares regression line .The three median line is a graphical method for fitting a line to data – being a graphical method it is quick and requires minimal computation. which is its strength – its other strength is that. being based on medians.

there should be:  No clear outliers This is not a requirement of the three median line. we say that we are extrapolating Predicting outside the range of data is called extrapolation For example. we would be interpolating because we are making predictions within the data – however. the three median line should be used in preference to the least squares line if there are any clear outliers However. we must be aware that.As mentioned in the example. the scale of measurement can be stretched or compressed There are many functions that can be used to transform data. even when there are no outliers. if we used the regression line to predict the carbon monoxide level for a traffic volume of 50 cars per hour. it is only luck if a data set reveals all its hidden information when analyzed in the form in which it was initially gathered and/or reported . for the least squares line. if we tried to use the regression line plotted on the scatterplot opposite to predict the carbon monoxide level for a traffic volume of 200 cars per hour. put one in each of the outer regions When fitting a three median or least squares regression line. because it is less influenced by outliers This is because it is based or medians rather than means Thus. the equation we have found only applies to the range of data values used to derive the equation Thus we are safe using the line to make predictions within this data range – this is called interpolation Predicting within the range of data is called interpolation However. we would be extrapolating as we are making a prediction outside the data Extrapolation is a less reliable process than interpolation because you are going beyond your original data There are methods for fitting curves to non linear regressions The method of dealing with a non linear relationship favoured in practice is to apply a mathematical function to one of the variable so that the relationship between the variables becomes closer to a straight line By appropriate choice of the function. it is not always possible to divide points into three equal sized groups If there is one extra point put it in the middle region If there are two extra points. strictly speaking. the three median line is sometimes used in preference to the least squares line as it is a quick and easy graphical technique which requires minimal computation When using a regression line to make predictions. we must be extremely careful about how much faith we put into predictions made outside the data range. as we have no way of knowing whether or not the equation we have derived applies When we make predictions outside of the data range. but predominately three are used:  The square transformation  The log transformation  The reciprocal transformation Generally. the key requirements are that the  Variables are numerical  The relationship is linear In addition.

and values between 0 and 1 become negative – large values are reduced much than small values – the effect of the log transformation is to compress the values The reciprocal transformation again reduces all values greater than 1.   The square transformation has the effect of decreasing values less than 1. and increasing values greater than 1 – large values are increased the most – that is. log and reciprocal transformations can be applied to the x axis with the following effects: . the effect of the square transformation is to stretch the data values The log transformation reduces all values.larger values are reduced much more than small values – the effect of the reciprocal transformation is to compress the large values to an even greater extent than the log transformation Transforming the x axis will have the effect of moving the x values on the plot horizontally and leave the y values unaltered – the square.

the square. log and reciprocal transformations can be applied to the y axis with the following effects: .Transforming the y axis will have the effect of moving the y values on the plot vertically. and leave the x values unaltered .

in order to evaluate the linearity of the transformed relationship  The value of the coefficient of determination (r2): a higher value indicates a better fit .The advantages of having alternatives to linearise is that in practice. we can always try each of them to see which gives us the best result – how do we desire which transformation is the best though? The best transformation is the one that results in the best linear model To choose the best linear model we will consider for each transformation applied:  The residual plot.

.

1 & 4. their magnitudes add to 180° Other pairs of supplementary angles are 2 & 3. and 6 and 8 Angles 1 and 2 are called supplementary i. 4 & 3. 6 & 7. 5 & 6.e. 5 and 7.Trigonometry & Geometry: Chapter 12 – 17             Angles 4 and 6 are called alternate angles Angles 5 and 3 are called alternate angles Angles 2 and 6 are called corresponding angles Angles 1 and 5 are called corresponding angles Angles 4 and 8 are called corresponding angles Angles 3 and 7 are called corresponding angles Angles 3 and 6 are called cointerior angles Angles 4 and 5 are called cointerior angles Angles 1 and 3 are called vertically opposite angles and are of equal magnitude Other pairs of vertically opposite angles are 2 and 4. the angles opposite each of the equal sides are equal . corresponding angles are of equal magnitude. 8 & 7. alternate angles are of equal magnitude and cointerior angles are supplementary Converse results also hold:  If corresponding angles are equal then l1 is parallel to l2  If alternate angles are equal then l1 is parallel to l2  If cointerior angles are supplementary then l1 is parallel to l2 a°. & 5 & 8 When lines l1 and l2 are parallel. b° and c° are the magnitudes of the interior angles of the triangle ABC d° is the magnitude of an exterior angle at C  the sum of the magnitudes of the interior angles of a triangle is equal to 180°: a° + b° + c° = 180°  b° + a° = d° the magnitude of an exterior angle is equal to the sum of the magnitudes of the two opposite interior angles   a triangle is said to be equilateral if all its sides are of the same length: AB = BC = CA the angles of an equilateral triangle are all of magnitude 60°  the bisector of each of the angles of an equilateral triangle meets the opposite side at right angles and passes through the midpoint of that side  a triangle is said to be isosceles if it has two sides of equal length – if a triangle is isosceles.

as shown in the diagram. b and c.  the sum of the magnitudes of the exterior angles of a triangle is equal to 360°: e° + d° + f° = 360° a triangle is said to be a right angled triangle if it has one angle of magnitude 90° A regular polygon has all sides of equal length and all angles of equal magnitude A polygon with n sides can be divided into n triangles – the first three polygons below are regular polygons The angle sum of interior angles of an n-sided convex polygon is given by the formula: The results hold for any convex polygon – convex means that a line you draw from any vertex to another vertex lies inside the polygon The magnitude of each of the interior angles of an n sided polygon is given by: The angle bisector of a regular polygon meet at a point O For a regular polygon a circle can be drawn with centre O on which all the vertices lie The sum of the angles at O of a regular polygon is 360° The sum of the exterior angles of a regular polygon is 360° Pythagoras’ theorem states that for a right angled triangle ABC with side lengths a. The sum of the areas of the two smaller squares is equal to the area of the square on the longest side (hypotenuse) . a2 + b2 = c2 Pythagoras’ theorem can be illustrated by the diagram shown here.

A corresponds to X. B corresponds to Y.b. C corresponds to Z .A triple of natural numbers (a. but is similar to rectangle 3: So for two rectangles to be similar. i.c) is called a Pythagorean triad if: c2 = a2 + b2 The table presents the first six such primitive triples – the adjective primitive indicates that the highest common factor of the three numbers is 1 Two objects are defined to be similar if they have the same shape by not the same size Any two circles are similar to each other Any two squares are similar to each other It is not true that any rectangle is similar to any other rectangle Rectangle 1 is not similar to rectangle 2.e. their corresponding sides must be in the same ratio Two triangles are similar if one of the following conditions holds: Corresponding angles in the triangle are equal Corresponding sides are in the same ratio Where k is the scale factor Two pairs of corresponding sides have the same ratio and the included angles are equal If a triangle ABC is similar to triangle XYZ this can be written symbolically as The triangles are named so that angles of equal magnitude hold the same position.

cylinders. m3 etc. rectangular prisms and triangular prisms The volume of a prism can be found by using its cross sectional area Answers will be in cubic units.e. i. cm3.A prism is a solid which has a constant cross section – examples are cubes. The formulas for determining the volumes of some standard prisms: The formula for finding the volume of a right pyramid can be stated as: The term right in this context means that the apex of the pyramid is directly over the centre of the base The formula for finding the volume of a cone can be stated as: . mm3.

new shapes can be made – the volumes of these can be found by summing the volumes of the component solids The surface area of a solid can be found by calculating the totaling the area of each of its surfaces – the net of the cylinder in the diagram demonstrates how this can be done Here are some more formulas for the surface areas in some solids: .The formula for the volume of a sphere is: Where r is the radius of the sphere Using the shapes above.

If two shapes are similar and the scale factor is k. then the area of the similar shape = k2 x area of original shape Two solids are considered to be similar if h have the same shape and the ratio of their corresponding linear dimensions is equal Some examples of similar volume and the ratio of their areas: .

then the volume of the similar solid = k3 x volume of original solid The unit circle is a circle of radius 1 with centre at the origin Sine and cosine may be defined for any angle through the unit circle .If two solids are similar and the scale factor is k.

e.The sine rule is used to find unknown quantities in a triangle when one of the following situations arises:  one side and two angles are given  two sides and a non included angle are given Interior angles are denoted by upper case letters and the length of the side opposite angle is denoted by the corresponding lower case letter The magnitude of angle BAC is denoted by A The length of side BC is denoted by a The sine rule The sine rule states that for a triangle ABC: The cosine rule is used to find unknown quantities in a triangle when one of the following situations arises:  two sides and an included angle are given  three sides are given The symmetrical results also hold. i.2ac cosB  c2 = a2 + b2 .2ab cosC It is known that the area of a triangle is given by the formula: .:  b2 = a2 + c2 .

Heron’s formula gives a way of determining the area of a triangle given the lengths of three sides The angle of elevation is the angle between the horizontal and a direction above the horizontal The angle of depression is the angle between the horizontal and a direction below the horizontal The three figure bearing (or compass bearing) is the direction measured clockwise from north Surveyors sometimes need to measure distances to inaccessible points. or measure lengths that are impossible to measure directly .triangulation involves using the theory of solving triangles developed in this module Problems in three dimensions are solved by picking out triangles from a main figure and finding lengths and angles through these triangles This diagram shows a hill over 200 m high rising from sea level – B is a point on the hill 50 m above sea level – a line drawn through b passes through all other points that are 50 m above sea level – this is called a 50 m contour line .

C. first determine from the diagram the horizontal distance B’C’ We already know that the vertical distance is 50 m as 100 – 50 = 50 Pretend that the horizontal distance between B and C is 80 m The triangle BCH below can be used to find the distance between B and C and the average slope between them The average slope = And Which implies = 32.Similarly. D and E mark levels of the 100 m. 150 m and 200 m contour lines respectively – imagine that the contour lines are all painted black and then projected onto the base as shown This is called a contour map Not that the map does not give the actual distance between B and C but gives the horizontal distance In order to find the distance between B and C.00° correct to 2 decimal places The angle of elevation of C from B is 32° to the nearest degree A cross sectional profile can be drawn from a contour map for a given cross section AB This is illustrated below: .