DataAnal V5.5 ============= 1. Introduction =============== DataAnal is a program for analysing data (no surprise there). At the very least, DataAnal will find the mean, median and standard deviation of a list of numbers. It will also draw histograms, plot graphs, and compare your data to mathematical probability distributions. DataAnal can read data in two forms: (a) A list of single numbers. (b) A list of coordinate pairs. Data is read from ASCII text files, and can therefore by typed up and edited with an standard text editor, such as `ED'. 2. Instructions =============== 2.1 Entering/Loading Data ------------------------- When the program is first run, you have a choice to load in the data from a disk file, or to enter it from the keyboard. Choose the appropriate item from the menu, using the right mouse button. To enter data from the keyboard, you must specify how many data items you have. Then just type in each number, and press return. If you are entering coordinate pairs, you must separate the two numbers with a comma. You can also enter a note corresponding to each item of data. This is done by typing a ' after the data, and then entering your note. This is explained in more detail later. 2.2 The Main Screen ------------------- The Main Option Screen lists a portion of the data list you have just typed in or loaded. On the right hand side there is some information about the current file, and the current status of DataAnal. The portion of data is divided up into five columns. The first column shows the item number of the data. Click on an item number to centre the data-list on that item. The second column (M) shows whether or not an Item has been `marked'. When results are calculated, only the `marked' items will be taken into account. This means it is possible to perform calculations on a certain portion of the data only. Marked Items are indicated by a "*". To mark/unmark an item, click in this column. The third and fourth columns (Set X and Set Y) show the data itself. If coordinate pairs were typed (or loaded) in then Set X and Set Y will contain numbers. Otherwise, only Set X will. The final column shows any notes you have added to the data-list. Notice one item in the data-list is in a darker colour. This is the one that will be amended if the `Amend' option is selected. The File Info on the right tells you how long the current file is (if it was loaded from disk), and how many items there are. Also written here is something like this:- Operator: Ignore No. Sets: 1 or 2 X Func : Linear Y Func : Linear `Operator' refers to which operator has been selected from the `Mark' menu. This is explained later. `No. Sets :' Tells you how many sets there are. If `Set Y' contains numbers, then this will be `2'. `Set X: Linear' (and `Set Y: Linear' if it is there) refer to the function that is currently operating on each data set. This is selected from the `Func' Menu. 2.3 Menus --------- Functions are generally selected using pull-down menus. Most functions are relatively straight forward... SAVE will save the current data-list. NEW will clear the list and return you to the opening screen. The PAGE UP, PAGE DOWN, GOTO FIRST and GOTO LAST enable you to list a different portion of the data. GOTO #N will allow you to choose a certain item number; the data-list will then be centred on this item. AMEND allows you to enter a new value of a specific item. The item that is currently shown in white will be amended. To amend a coordinate pair, you must enter both coordinates (x then y), separated by a comma. You can also write a short note for each item. This is preceded by a '. For example, you might type in: 1,3 'This is my note for a coordinate 27 'This is my note for a single item As well as marking items manually, by clicking on the `M' Column as described above, you can also mark a whole section of the data at a time. MARK ALL and SWAP ALL are self-explanatory. Here are the more interesting ones... If you choose MARK ABOVE you will be asked to enter a number. All items whose data is above this number will be "marked". Similarly for MARK BELOW. With MARK RANGE, a continuous portion of the data can be selected (eg, all items from item 3 to item 7). You can also perform AND, OR and XOR operations with the previous `mark' status. To do this, select the appropriate operator from the `Mark' menu. For example, suppose you wanted to select all items that were either above 20 or below 10. You would first choose MARK ABOVE and type in 20. Then select OR from the list of OPERATORs so that the results from both search will combined. Finally select MARK BELOW and type in 10. If you want the previous select status to be ignored, choose the `Ignore' operator. This is the default. The options in the RESULTS menu are examined in section 2.4 The next menu is the FUNCTION menu. This is used to perform a function on every item of data in Set X and/or Set Y. The default is `Linear' which simply means no function is in operation. Here is a list of the functions and the result on the data set. Function Effect -------- ------ Linear Nothing Square Take the square (x^2) Square Root Take the square root Inverse Take the reciprocal (1/x) Ln Take the natural Logarithm (ie, to base e) Exp Take the exponent (ie, inverse of log) All calculations will then apply to the values of the functions of x, not the original value itself. For example... Linear Square ------ ------ Item1: 2 4 Item2: 3 9 Item3: 6 36 Item4: 5 25 Mean: 4 18.5 The final menu is the SETS menu. These options will do things to your entire data set(s). `Swap X,Y' simply swaps Set X and Set Y around. `Kill Set Y' will remove the numbers in `Set Y', so that only one data-set (Set X) is left. `Sets 1<->2' will change your data list from one to two sets (or back from two to one). This works as follows: When a single set is converted to two sets, then the number of times each element occurs becomes the corresponding second set. Here is an example: Set X Set X Set Y 2 2 2 (ie `2' occurs twice) 2 -----> 3 1 (ie `3' occurs once) 3 4 0 (ie `4' does not occur at all) 5 5 3 (ie `5' occurs three times) 5 5 Note the order of Set X does not matter. If Set X had contained 5,5,5,3,2,2 then the result would have been the same. It is important to realise therefore, than using this option will cause the order of the data to be lost. Also, if any of your data-items had `notes', then these will be lost too. Selecting this option again will convert the two sets back to one set. 2.4 Results ----------- Note: Results are only calculated for those items that have been `marked'. Initially, all the items are `marked'. This first item in this menu is Print. You can set the print device to a the printer (ie, PRT:) or filename (eg, DF0:myfile). All results will then be outputed to your chosen print device/file. To cancel this, select `Cancel' from the same menu. The results will apply as follows: If there are two data sets, then the item in Set Y is assumed to mean to frequency with which the corresponding item in set X occured. Here is an example: Set X Set Y 1 2 } ---------> 1, 1, 2, 2, 3 The results refer to this 2 2 } list of numbers. Eg, the 3 1 } average (mean) is 1.8 LIST will simply give a list of results. The mean, standard deviation, median, minimum value, maximum value, and mode (if appropriate) are shown. Note that the results only apply to items that have been marked. HISTOGRAM will draw a histogram showing the frequency with which each number occured. This is only applicable to integers, but for non-integers, the numbers are rounded first. You can choose the character with which the histogram is to be drawn (usually # or *). You can also make the histogram bigger by, for example, entering `##' instead of `#'. Here is an example: Set X or Set X Set Y Item1: 7 4 1 Item2: 6 5 0 Item3: 7 6 2 Item4: 4 7 3 Item5: 7 Item6: 6 would produce a histogram like one of these: Characters=`#' Characters=`##' Characters=`*' 4:#1 4:##1 4:*1 5:0 5:0 5:0 6:##2 6:####2 6:**2 7:###3 7:######3 7:***3 PROBABILITY will attempt to fit your data to a probability distribution. In this version, the following distributions are available: - Binomial. - Pascal (or Negative Binomial). - Normal. Both the actual probabilities (ie, according to your data), and the theoretical probabilities (ie, according to the distribution itself) are shown for each value of data. P(x=a) means the probability that x equals a. P(x<=a) means the probability that x is less then or equal to a. DISTRIBUTION will also attempt to fit your data to a probability distribution, but will draw a graph of the distribution, rather than a table of probabilities. The continuous line is the theoretical distribution, and the solid vertical lines represent the frequencies with which each data value occured. For each distribution, the defining coefficients are also displayed. Note: You may not be able to fit your data to ALL of the distributions. The computer will tell you if it is impossible. The Normal Distribution should work for all data, however. PLOT Y vs X will only work if there are two sets of data. If you select this option, a graph will be plotted of y against x. Before the graph is plotted, you have a choice of two options: Best Fit Line enables you to choose whether you want a line of best fit drawn on the graph. Lower-Left Origin enables you to decide whether the lower-left corner of the graph will be the origin(0,0) or the smallest value of data. `m=' and `c=' give the gradient and y-intercept respectively, of the best-fit line. It is also possible to plot a graph of say, Ln Y against X. To do this, just choose `Set Y/Ln' from the Functions menu, before plotting. 3. Further Notes ================ 3.2 DataAnal_Files Directory ---------------------------- (a) When you opt to load a data file from disk, DataAnal will look for a directory called `DataAnal_Files/'. If no such directory exists, DataAnal will instead look in the current directory. If you decide to use DataAnal a lot (!), it would therefore be a good idea to keep all your data-lists in the `DataAnal_Files/' directory. (b) By way of an example to show how DataAnal could be used, I have included a file called `Presidents.DA'. It is in the `DataAnal_Files' directory. This contains the ages of all the Presidents of the United States when they took office. The item's notes are the corresponding Presidents' names. You should therefore be able to work out the average age of American presidents, and find out who was the youngest etc. Sorry I couldn't think of anything else more interesting to include instead! 3.2 Format of Data Files ------------------------ To make life simple, the data is saved as simple ASCII text files. These can be edited with any text editor. The items must be on separate lines. Lines that start with letters are ignored. Example... 24 'hello1 hello2 26 'hello3 26 'these are notes 22 'this is another note DataAnal would read: 24,26,26,22. The corresponding notes would be`hello1',`hello3',`these are notes', and `this is another note'. `Hello2' would be ignored. Coordinate pairs are read and saved in this format: 1.2, 2.4 'note1 1.6, 3.1 'note2 2,4, 3.7 'note3 2.8, 4.5 'note4 3.3 Author ---------- DataAnal V5.5 (c) Steve P, 1989-1993 Steven Pugh 10 Wellington Close Sandhurst Green Sundorne Shrewsbury Shropshire SY1 4SP Write to me if you any suggestions for improvements, or if you can think of any interesting uses for this program. DataAnal is freely distributable provided this document, and the file `Presidents.DA' is included in all copies.