Research Methods Workshop Introduction to EViews

Transcription

1 Research Methods Workshop Introduction to EViews

2 Today s Workshop Brief Introduction to EViews and Basic Operations Descriptive Statistics with EViews Group Statistics with Eviews Note that the data we will use today can be downloaded from: mybu/unit Materials/Workshops 3&4: SPSS-Eviews/Eviews Today we will mainly use Quarterly Economic Index Monthly Financial Index

3 Brief Introduction to EViews and Basic Operations Workfiles, objects and their basic operations Importing data Structure your data Data operation Graphical skill Fundamental operation and function

4 Descriptive Statistics Different ways of obtaining descriptive statistics for your data Interpreting descriptive statistics Histogram Series Statistics

5 Group Statistics Correlation Cointegration Simple linear regression model Interpreting the results Introduction to multiple linear regression modelling

6 What is Eviews (Econometrics Views)? EViews is an easy-to-use statistical, econometric, and economic modeling package, which provides data analysis, regression and forecasting tools: Useful for many different analyses Very user-friendly (like Windows) Excellent help function In this introduction workshop, we will introduce two different ways to work in EViews: Graphical user interface (using mouse and menus/dialogs). Single commands (using the command window).

7 Getting Start EViews Desktop Main Menu Command Window Object Window/ Work Area Note: Path/Database/Workfile can be changed by doubleclicking in each. Path/directory Database Workfile

8 Workfiles and Objects EViews does NOT open up with a blank generic document (unlike Word, Excel, etc.). EViews documents ( workfiles ) need to be created and are not generic (they will contain information about your data, etc.). EViews is an object - oriented program. Objects are collections of information related to a particular analysis (series, groups, equations, graphs, tables). Workfiles are holders of these objects, it is the central place to keep all of your work.

9 Eviews Workfile Workfile title bar Workfile tool bar Workfile Window Workfile: Contains at least one page Each page contains a list of objects on that page

10 Eviews Workfile Structure of the workfile The data in this example is dated and has monthly frequency covering the period from 1977 to Name of the workfile Monthly financial index in this example) Range: shows the entire range of the data in the workfile. Here the range is from M to M Sample: This is the part of data we are currently working with. In this example, the sample runs from Q to Q

11 Example Eviews Workfile and Objects This screenshot shows a list of Objects in the example workfile. It is color-coded by Object type: Yellow icons are data objects Blue icons are estimation objects Green icons are view objects (tables, graphs, etc ) Double clicking on one of these Object icons will open it up. Each Object has its own menu. Once an object is open, the menus in EViews change to represent the features available for that object.

12 Series, Groups and Equations are the most common objects in EViews. Symbols of Different Types of Objects

13 The Object Window Main menu Workfile Toolbar Object Toolbar (in this example, series toolbar) Object Window (in this example, series window)

14 The Series Object This is the main data object. sp500 - has a yellow icon with a little line graph in it. It contains one column of data. Opening a series will reveal a spreadsheet view with a single column showing the data in the series.

15 The Group Object This is a collection of series objects. group01 - has a yellow icon with a capital G. It contains multiple columns of data. Opening a group will display a spreadsheet view with multiple columns showing the data in each series in the group.

16 Workfiles Operations Any work in EViews is created in workfiles which are place-holders of EViews objects. Unlike other types of programs, an EViews workfile must be created. Every workfile contains one or more workfile pages, each with its own objects. To set-up a workfile, you can either: Create a blank workfile (this requires structuring); Create a workfile by reading from a foreign source.

17 Creating a Workfile Creating a blank workfile Main menu File New Workfile Data Structure (Frequency, start/end dates, etc.) The structure determines how many observations (i.e. rows) every series in the workfile page contains. There are three general types of workfile structure: Time Series Dated/Regular Frequency + Date Range Cross-Sectional Undated/Unstructured Range of observations Balanced Panel (panel data) Save & Name

18 Creating a Workfile Area Use the Workfile structure type pull down menu to specify the general structure type. Area 2. Use the Date specification area to specify the frequency and the start date/time and end date/time. Area 3. Enter a specific workfile name and first page for the workfile

19 Creating a Dated Workfile: Example 1 To create a workfile with monthly data from 1950 to 2012: 1. Click File New Workfile 2. The Workfile Create dialog box opens up. 3. Select Dated-regular frequency from the drop-down menu of Workfile structure type. 4. Select Monthly from the drop-down menu of Date specification. 5. Enter the start and end date of your data ( ).

20 Creating an Undated Workfile: Example 2 To create a workfile with GDP data for all 50 states (50 observations) 1. Click File New Workfile 2. The Workfile Create dialog box opens up. 3. Select Unstructured/Undated from the drop down menu of Workfile structure type. 4. Type the number of observations in Data Range (in this case 50).

21 Creating a Balanced Panel Workfile: Example 3 To create a Workfile Containing Quarterly Data from for 100 countries 1. Click File New Workfile 2. Select Balanced panel from the drop down menu of Workfile structure type. 3. Select Quarterly from the drop down menu of Panel specification. 4. Enter the start and end date of your data (here ). 5. Enter the number of cross sections (here 100 countries).

22 Example 1 Creating Workfiles Using Commands Enter the following command into the command window, then press Enter wfcreate(wf=monthly, page=dated) m The new workfile is named MONTHLY containing a single page named DATED, structured as a monthly workfile from Example 2 Enter the following command into the command window, then press Enter wfcreate(wf=undated, page=state) u 50 The new workfile is named UNDATED containing a single page named STATE, structured as undated workfile with 50 observations. Example 3 Enter the following command into the command window, then press Enter wfcreate(wf=panel, page=country) q The new workfile is named PANEL containing a single page named COUNTRY structured as a panel workfile with 100 cross-sections of quarterly data from

23 Creating Workfiles from foreign data source EViews can create a new workfile by opening data from a variety of formats (Excel, HTML, CSV, ASCII, RATS, Stata, SPSS, SAS, etc ). EViews automatically recognizes the format and structure of files. Simply open the file in EViews and let the application do the rest. There are a number of ways to open a file in EViews: a) Drag/drop a file in the main EViews window (this is the easiest method). b) Click File Open Foreign Data as Workfile. c) Copy/Paste as New Workfile in EViews desktop.

24 Dragging/Dropping Data: Example 1 1. The easiest way to create a workfile is to select a file and drag/drop it onto the EViews desktop. 2. A + (plus) Copy sign appears when the file is in the appropriate area. 3. The Excel Read dialog appears. 4. In most cases (including this one), clicking Finish on this dialog will load the data. 5. EViews scans the Excel file to determine the structure. Note: If your data has lines of text before the column of data (showing data sources, and other info), you can tell EViews to skip these lines in the Column headers section which appears if you hit Next instead of Finish.

25 Importing Data: Example 2 1. Click File Open Foreign Data as Workfile. 2. Select the directory where data is stored and open file. 3. The Excel Read dialogue appears again. 4. As in the previous examples, click Finish to load the data.

26 Copy/Paste: Example 3 1.Copy the source file (in this case the range of Excel rows). 2.Right-click anywhere on the EViews desktop. 3.Select Paste as New Workfile. 4.As in the previous examples, the Excel Read dialog appears; click Finish to load the data.

27 Importing Data by Command You can just as easily open a foreign source file in EViews by typing in the command window. 1. Enter the following command into the command window, then press Enter wfopen(page=test) c:\test.xls This creates a new workfile page named test. The second part of the command instructs Eviews where the file is located. 2. The Excel Read dialog appears again, then click Finish to load the data.

28 Object Operation Series and Groups are the most important Data Objects in Eviews. The actual numeric values of your data are held in the Series Object. A collection of Series (multiple data columns) comprises a Group Object. This section demonstrates basic work with Series and Groups: Creating a Series Bringing data into EViews Creating New Series from Existing Series Editing, documenting and displaying a Series Creating Groups and working with Groups

29 Creating a New Series Example 1 1. Select Object New Object from the main menu. 2. Click the Series option, name it (in this case series1). 3. Click OK. Example 2 Type the following command in the command window, then press Enter series series2 Note that the new series created by either method will have the structure of the workfile and NA in all entries.

30 Getting Data into an Existing Workfile Example 1 1.Open one of the new series with empty (NA) values we created above. 2.Click the Edit+/- button to turn on editing mode. 3.Start entering in data. Example 2 1.Open the data in Excel (or other applicable applications ). 2.Highlight the series you wish to bring into Eviews and select copy (note that you do not need to copy dates since the Eviews workfile is already structured by date). 3.In EViews workfile click Quick Empty Group (Edit Series). 4.A window opens up with a blank spreadsheet and dates on the side. Place the cursor on the upper-left cell, right-hand-click, and choose Paste.

31 Getting Data into an Existing Workfile Example 3 (from foreign files) 1.On the menu bar, select File (or Proc) Import Import from file. A standard File Open dialog box appears which allows you to locate the file. 2.Click Open once you locate the file. The Excel Read dialog box opens up prompting for additional information for the import procedure. 3.Click Finish to load the data.

32 Transforming Existing Series In order to create a series y=log(sp500) Note that in Eviews, log(x) is the natural logarithm (not of base 10) Example 1 In the command window, type the following command then press Enter. Example 2 series y=log(sp500) or genr y=log(sp500) 1.click Quick Generate Series from the menu toolbar. 2.The window Generate Series by Equation opens. Type your data transformation here (in this case, x=log(sp500)). 3.Click OK.

33 Groups Groups help you work with multiple series. A Group is a list of series names (and potentially mathematical expressions) that provides access to all the data in that list. Once you create a Group Object, you can use the group name in many places to refer to all the series contained in that group. A few features of groups: A group is a live feed and is NOT a copy of each individual series. This means that if the data in one of the series changes, these changes will also be reflected in the group containing the series. If a series is deleted from a workfile, the series identifier will be maintained. In the group spreadsheet the deleted series will contain NA values. Renaming a series changes the reference in every group containing the series.

34 Creating Groups Example 1 1. Select Object New Object from the main menu. 2. The New Object box opens up. Select the Group option. 3. You can name your group under the section Name for object (in this case, we named it group01), then press OK. 4. The Series List window appears. Enter the series names you wish to include in the group (separated by spaces). Example 2 1.Select Quick Show from the main menu (or Show from the workfile menu). 2.The Show window appears. Type here the names of the series you wish to include in the group 3.Press OK. 4.Click Name button to save the group. 5.The Object Name box opens up, name the group here.

35 Creating Groups Example 3 1. Highlight the series you wish to group together. 2. Right-click and select: Open as Group. 3. As before, Click Name button to save the group. 4. The Object Name box opens up, name the group. Example 4 Type the following command in the command window, then press Enter. group group03 tbill1y tbill3m tbill6m This produces a group named group03, which includes series tbill1y, tbill3m and tbill6m.

36 Graphing in EViews Graphing is an important part of the process of data analysis and presentation. EViews provides a powerful, user-friendly, full-featured set of tools that will aid you in graphically displaying your information. Today we focus only on basic graphing in EViews. The main topics are: Graph Options and Graph Objects Graphing Single Series Graphing Multiple Series

37 Creating a Graph 1. Open series SP Click View Graph. 3. This brings up the Graph Option dialog box. Under Option Pages select Basic type page. 4. Under the Graph type section, select Basic graph from the General category. 5. Select Line & Symbol from the Specific category. 6. Leave Details settings as specified by default options (and as shown here). 7. Click OK. 8. Click Freeze to save the graph by creating a new graph object 9. Click Name to save this new graph object 10. The Object Name box opens up, name the graph.

38 Graph Options Menu The Graph Options menu is broken into several categories: 1. Option Pages: This is used for customization. 2. Graph Type: this specifies the type of graph you wish to display. It has two categories: General and Specific General this combo box has two options: o Basic graph: shows basic graph of data. o Categorical graph: shows data divided into categories defined by factor variables Specific: offers a list of the graph types available for use. Note that there are different options available for series and groups. Series Groups

39 Graph Options Menu 3. Details: this category specifies additional details, such as: Graph Data Orientation Axis Borders Multiple Series Graph Data specifies the data to be used in observation graphs o Raw data is the default setting. Every observation is plotted. o Summary Statistics allows you to compute summary statistics (means, medians, etc). For a single series, the summary statistics shows a single data point. For single series, you may want to leave the setting at Raw data. The Graph data option is more useful with multiple series (Groups)

40 Graph Options Menu Orientation allows you to choose between two options: o Normal observations are along the horizontal axis. o Rotated observations are along the vertical axis. Please note that for three types of graph Distribution, Quantile-Quantile or Seasonal Graphs different options appear instead of Orientation. odistribution Graphs the Orientation option changes to Distribution with the drop-down menu shown here.

41 Graph Options Menu oquantile-quantile Graphs the Orientation option changes to Q-Q graph option with the drop-down menu shown here. oseasonal Graphs the Orientation option changes to Seasonal type option with the drop-down menu shown here.

42 Graph Options Menu Axis Borders allows you select a distribution graph to display along the axis of your graphs. o By default no axis are displayed (default None). Multiple Series this allows you to handle multiple series (group data). o If you have a single series, this menu option is unavailable. o For multiple series, it shows the options displayed here which ask you to specify how you want to display your multipleseries.

43 Graph Object When creating graphs by clicking on the Graph view from the View menu of a series or group, remember that what you create are simply graph views. These graph views are transitory: while you can customize them, once you close the series or group object, most of these settings will be lost. You may save a graph by freezing it. This creates a graph object. Freezing a view (graph object) creates a snapshot of the current graph view, allowing you to create permanent customization of the output.

44 Single Series Simple Graphs Example 1: Line & Symbol The simplest type of graph is the Line & Symbol. The graph shows data (as symbols, line or both) against observation identifiers. Line & Symbol graph: 1. Open series tbill3m. 2. Click View Graph. 3. This brings up the Graph Option dialog box. Under Option Pages select Basic type page. 4. Under the Graph type section, select Basic graph from the General category. 5. Select Line & Symbol from the Specific category. 6. Leave Details settings as specified by default options (and as shown here). 7. Click OK.. 44

45 Single Series Simple Graphs Example 2: Dot Plot The dot plot is a symbol-only version of the Line & Symbol graph. The graph uses circles to represent the value of each observation. Dot Plots: 1. Open series tbill3m. 2. Follow steps 2-4 of Example Select Dot Plot from the Specific category. 4. Leave Details settings as specified by default options (make sure Raw Data is selected under Details/Graph Data). 5. Click OK. The graph is shown here.. 45

46 Single Series Simple Graphs Example 3: Bar Graph Bar Graphs use bars to represent the value of each observation. Bar Graphs are effective when there are few observations: for large numbers, Bar Graphs and Area Graphs are almost indistinguishable from each other. Bar Graphs: 1. Open series tbill3m.(choose sample1977m01 to 1978m01) 2. Follow steps 2-4 shown in Example Select Bar (from Graph type/specific). 4. Leave Details settings as specified by default options (make sure Raw Data is selected under Details/Graph Data). 5. Click OK..

47 Single Series Simple Graphs Example 4: Spike Plots Spike Plots use a thin bar to represent the value of each observation. Spike Plots are essentially Bar Graphs (with thin bars) and can be used to display a moderate amount of observations. Spike Plots: 1. Open series tbill3m. Remember the sample is still set in 1977m01 to 1978m01) 2. Follow steps 2-4 shown in Example Select Spike (from Graph type/specific). 4. Leave Details settings as specified by default options (make sure Orientation is set on Normal and Axis borders on None). 5. Click OK.

48 Single Series Simple Graphs Example 5: Area Plots An Area Graph is a line graph with the area underneath the line filled in. Area Graphs: 1. Click on the Panel page. 2. Let s first create a new series by subtracting its own mean from the tbill3m series. For this, type in the command window: series 3. After creating it, now open the series newseries. Follow steps 2-4 shown in Example Select Area (from Graph type/specific). 5. Click OK.

49 Multiple Series Simple Graphs All types of graphs available for a single series are also available for graphing series in a group. Multiple Series have all the graph options of Single series along with a few additional ones. One important difference is that multiple series graphing has one additional option to choose that all plots in single graph or in multiple graphs separately. If you open a group of series, the following graph types appear: 49

50 Multiple Series Simple Graphs Example 1: Line & Symbol -- One Graph Line & Symbol One Graph: 1. Open group Click View Graph. 3. This brings up the Graph Option dialog box. The default is set on Basic type page; keep the default setting. 4. Under the Graph type section, select Basic graph from the General category. 5. Select Line & Symbol from the Specific category. 6. Notice that now under Details/Multiple Series you have a few options. Let s select Single Graph. 7. Click OK.

51 Multiple Series Simple Graphs Example 2: Dot Plot Multiple Graphs You can also plot each series separately if you wish. Let s demonstrate this by graphing a Dot Plot. Dot Plot with Multiple Graphs: 1. Follow steps 1-4 in example Select Dot Plot (from Graph type/specific). 3. Under Multiple Series select Multiple graphs. 4. Click OK. The graph is shown here. Notice that in contrast to the Single Graph setting where each series is plotted on the same scale, here each graph has its own scale.. 51

52 Multiple Series Simple Graphs Example 3: Area Plot Stack in One Graph You can also stack your series in a single graph. This sums the values of each series vertically. Let s demonstrate this by graphing an Area Plot. Area Plot Stack in One Graph: 1. Follow steps 1-4 in example Select Area (from Graph type/specific). 3. Under Multiple Series select Stack in single graph. 4. Click OK. The graph is shown here. Notice that here the first series (tbill1y) is plotted in the usual way. The second series (tbill3m) is the sum of the first and the second series. The third series (tnote6m) is the sum of the first three series, etc..

53 Multiple Series Simple Graphs Example 4: Bar Graph Summary Statistics With multiple series it makes sense to take a look at summary statistics (means, median etc.) Let s plot the means of each series using a Bar Graph. Bar Graph Means: 1. Follow steps 1-4 in example Select Bar (from Graph type/specific). 3. Under Graph data select Means. 4. Click OK. The graph is shown here. Each bar represents the mean of each series.

54 Descriptive Statistics in EViews View menu: The main data analysis item in EViews is the View menu. You can view all available actions for a series from this menu. 1. Open one series 2. Click on the View menu item. A dropdown menu appears with a number of options grouped in 4 sections. The first section/block lists views that display the data series. The second bloc provides general statistics. The third block provides general statistics for time series. The fourth block allows you to modify/display the series labels.

55 Histogram and Stats Basic statistical summaries (including histograms) can be found in EViews under the View Descriptive Statistics & Tests Descriptive Statistics: Histogram 1. Open one series 2. Click View Descriptive Statistics & Tests Histogram and Stats.

56 Histogram and Stats The table of the right side will show if you click View Descriptive Statistics & Tests Stats Table.(this is easy to copy and paste into a word processor or spreadsheet program) 1. The top of the statistics panel shows the name of the series, the sample(only for Histogram and Stats ), and the number of observations (bottom of Stats Table). 2. In the middle, EViews summarizes the main statistics for the GDP series. (note: skewness measures the degree of asymmetry in data, where as kurtosis measures the presence of fat tails ). 3. The bottom of the panel shows the Jarque-Bera statistic and its associated probability and tests whether the data is drawn from a normal distribution. Since here, the p-value is , it is extremely unlikely that the data follows a normal distribution.

57 One-Way Tabulation If you want to see the entire distribution of the data, you can do so by using One-Way Tabulation. 1. First, make sure you include all observations in the workfile by typing in the command window: 2. Open series GDP 3. Click View One-Way Tabulation. The Tabulate Series dialog box opens up. Let s uncheck the boxes under the section Group into bins if. 4. The tabulation is shown here. As you can see, EViews provides the counts, percentage counts and cumulative counts for each observation value.

58 View Menu for Groups View menu for Groups: 1. Highlight series g gdp and inv, right click and select Open as Group. 2. Click on the View menu item. A drop-down menu appears with a number of options grouped in 4 sections. The first section/block provides various ways of looking at the data in the group. The second bloc provides general statistics. The third block provides general statistics for time series. The fourth block allows you to modify/display the group labels.

59 Descriptive Statistics: Common / Individual Sample Basic statistical summaries of the series contained in one group can be found in EViews under the View Descriptive Statistics. 1. Open group Click View Descriptive Statistics Common / Individual Sample. General stats for all the series in the sample are shown here. Note that statistics are computed using a common sample. This means that if a series has missing observations, stats for all series are computed over the common sample for which all series have non-missing observations. You can also compute Descriptive Statistics for the series in one group over the sample of each individual series.

60 Statistical Analysis in EViews Data analysis is one the most basic but important tasks when working with series and groups of series. EViews offers an impressive toolkit when it comes to data analysis, which include descriptive statistics, statistical graphs, tests and procedures. Most of the data analysis functions can be found in the View and Proc menu items. The main statistical analysis tests EViews covered are summarized as below: Series General Statistics Simple Hypothesis Test Empirical Distribution Test Correlogram Unit Root Test Groups General Statistics Covariance Analysis Principle Components Correlogram Cross Correlation Cointegration Test Granger Causality

61 Running Regression in EViews The main topics include: Specifying and estimating an equation Equation Objects (saving, labeling, freezing) Equation Output: analyzing and Interpreting results Basic Multiple Regression Analysis

62 Equation Object Single equation regression estimation in EViews is performed using the Equation Object. There are a number of ways to create a simple OLS Equation Object: 1. From the Main menu, select Object New Object Equation. 2. From the Main menu, select Quick Estimate Equation. 3. On the command window type: ls

63 Equation Estimation Box In all cases after the previous step, the Equation Estimation box appears. You need to specify three things in this dialogue box: 1. The equation specification. 2. The estimation method. 3. The sample. Specify your equation either by: a) List b) Formula Specify your estimation method (more methods in dropdown menu) Specify your sample

64 Specifying an Equation You can create an Equation simply by selecting the series and opening them as Equation. 1. Select employment_retail, tnote10y and tnote2y by clicking on these series in the workfile (press CTRL to select multiple series). Notice that you need to select the independent variable first. 2. Right click and select Open as Equation.

65 Specifying an Equation In the Equation Estimation dialogue box, you can type in the Equation specification box to specify your equation (same as the previous example): employment_retail c tnote10y tnote2y Note that always dependent variable first, followed by list of independent variables and constant term c. OR in the command window by typing in the following command then press Enter: ls employment_retail c tnote10y tnote2y

66 Regression Output Top Panel All previous methods will finally give you the estimation output in equation object. Click the Name button to name and save the equation object. Freeze the results will save the equation output in a table which will never change. The Equation box has three main parts, which we will discuss in turn: 1. The top panel summarizes the input for the regression. 2. The middle panel summarizes information about regression coefficients. 3. The bottom panel provides summary statistics about the entire regression. Middle Panel Bottom Panel

67 Regression Output Element Dependent Variable Method Date/Time Sample Description Denotes the dependent variables. Denotes the method of estimation (least squares in this example). Shows the date and time when the regression was carried out. Shows the sample period over which the regression is carried out. Included Observation Shows the number of observations included in estimation. Coefficient Values measures the marginal contribution of independent variable to dependent variable. C is the estimated constant (or intercept) of the regression. Standard Errors Reports the standard errors of the coefficient estimates. The larger the standard errors, the more noisy the estimates. t-statistic Reports the t-statistics, computed by dividing coefficient estimates by their standard errors. Is used to test whether the coefficient in that row equals zero. Prob. (p-value) Reports probability of drawing a t-statistic as extreme as the one actually estimated. Is used to test whether the coefficient is equal to zero (against a

68 Regression Output Statistic R-squared Adjusted R- squared S.E. of regression Sum squared resid Log-likelihood F-statistic Prob(F-Static) Description Measures the success of the regression in predicting the values of depended variable. Adjusts for the number of independent regressors by penalizing R-squared for additional regressors. Is a summary measure based on estimated variance of the residuals. Reports the sum of squared residuals. The same as (S.E. of regression) 2 * (T-k-1), where T is the number of observations, k is the number of independent variables. Reports the log likelihood function evaluated at coefficient estimates assuming normally distributed errors. Tests whether all slope coefficients (excluding the constant) are zero. Reports the probability of drawing an F-statistics as the one estimated.

69 Summary of Frequently Used Commands show variable show variable_a variable_b series variablename = equation genr variablename = equation group groupname variable_a variable_b smpl n1 n2 OR n2 OR smpl n1 and n2 represent the start and end of range equation name.ls dependent c independent variables cor variable_a variable_b cov variable_a variable_b

70 Thanks for your attention! Website: microsites.bournemouth.ac.uk/mathssupport/

71

72 Fundamental Operation and Functions Except using Quick Generate Series from the main menu, EViews also has a number of functions/commands that allow you to draw from a variety of distributions. The most commonly used functions/commands are summarized below. (simply type the command in the command window then press Enter) Function/Command Description series y=nrnd or (y=@rnorm) Normal distribution (mean 0, st. dev. 1) series y=3+@sqr(4)*nrnd Normal distribution (mean 3, variance 4) series y=@rlognorm(1,4) Lognormal distribution (mean=1, st. dev 4) series y=@runif(1,3) Uniform distribution on (1,3) series y=rnd Uniform distribution on (0,1) series y=1+(3-1)* rnd series y=rndint(0,100) Uniform distribution on (1,3) (same Fills series y with random integers drawn randomly from 0 to 100

73 Time Series Functions Common Commands/Functions Type Function Description Lags & leads gdp(-4) gdp(2) Denotes the 4 th lag of the GDP series Denotes the 2 nd lead of the GDP series gdp (-1 to -4) Specifies all GDP lags from 1 to 4 gdp(to -5) OR gdp (0 to -5) Specifies all GDP lags from 0 to -5 series y series y =@lag((gdpinv)/gdp,4) d(gdp) = gdp - gdp(-1) Generates series y, as the 3 rd lag of GDP Generates series y, as the 4 th lag of the transformation (gdp-inv)/gdp Takes the first difference of the GDP series Differences Trend d(gdp,3) dlog(gdp) = log(gdp) 3rd order difference of GDP series Takes the first difference of log(gdp) series 4 th order difference of log(gdp) series Time trend increasing with each observation Quadratic time trend increasing with each observation

74 Time Series Functions Type Function Description calculates the one-period percent change in GDP (in percent) calculates the one-period percent change in GDP (in decimal) calculates the one-year percent change (in percent) calculates the one-year percent change (in decimal) Cumulative @cumbmean(x,s) Cumulative Sum of the values in X over sample s Backwards cumulative sum of the values of X over sample s beginning with the end of the sample Backwards cumulative mean(or standard deviation) of the values of X over sample s beginning with the end of the sample up to and including current observation Note that sample S is programming as for example 1980Q3 1984Q4, so 1980Q3 1981Q4 ).

75 Empirical Distribution Tests EViews can carry out a number of empirical distribution tests, which test whether the data generating process for our series comes from one of these distributions: Hypothesis Tests: Example 4 1. Open GDP series. Click View Descriptive Statistics & Tests Empirical Distribution Tests. 2. The EDF Test dialog box opens up. There are two tabs: Test Specification allows you to specify the parametric distribution against which you want to test the empirical distribution of the series. Estimation Options provides control over any iterative estimation performed. This tab is mostly used in those cases when there is a failure in the estimation process. Otherwise you won t need to use this particular tab. 3. Under the Parameters section you can specify the values of any known parameters. If you leave it blank, EViews will estimate these from your data. 4. Click OK.

76 Function Descriptive Statistics-Functions EViews has extensive built-in descriptive statistical functions and the common commands/functions are summarized as follows: series series series 1980m m12 ) series y=@median(x) series y=@vars(x) series y=@varp(x) or y=@var(x) series y=@stdev(x) or y=@stdevs(x) Description Computes the geometric average of X Creates a series where each observation is equal to the mean of X Creates a series where each observation is equal to the mean of X for the defined sample (1980m01 to 1990m12) Creates a series where each observation is equal to the median of X Computes the sample variance of X (adj. by n-1) Computes the population variance of X (adj. by n) Computes the sample st. dev. of X (adj. by n-1) series y=@stdevp(x) Computes the population st. dev. of X (adj. by n)

77 Covariance Analysis The Covariance Analysis view is a very useful tool to obtain different measures of association (covariance/correlation) for the series in a group. There are four general classes in EViews from which you can compute measures of association: ordinary (Pearson), ordinary uncentered, Spearman rank-order, and Kendall s tau-a and tau-b. 1. Open group Click View Covariance Analysis. The Covariance Analysis dialog box opens up. Under Method, click the drop-down menu choose the type of measure (let s select Ordinary here). 3. Click the Covariance box (the default). 4. Under Saved results basename, name your results so they are saved in the workfile (here cov1) 5. Click OK.