Description
Background
This project will give you a first glimpse of building a program with many parts that must be
integrated into whole package. The project will be conducted in three stages. This will give you a
sense of how real live projects are conducted. You will be working in pairs (called pair
programming), which has been found to maximize student learning.
The project will be to build an application that computes (simple) statistics for a sample of data.
In this case the package supports floating point data (actually stored as doubles to minimize
round-off error). The data will be entered into a data file and the user can direct the application to
manipulate the data and compute the statistics. For this version we will stick with very simple
statistics of the data set, the mean, median, variance, and standard deviation. For those who are
interested in extending this project it is also possible to learn how to plot distributions and
incorporate several statistical tests, but that is not part of the requirements. Useful resource at
http://www.chem.utoronto.ca/coursenotes/analsci/StatsTutorial/MeasMeanVar.html
Product Description
StatPac is a simple package that allows a user to capture a set of data, as real numbers in floating
point format, manage the data set (e.g. adding more data, editing data, etc.) in files, and then
having the program compute and update the statistics for that data set. The program can display
the data on the screen or save to a formatted file for printing. It will also provide reports on the
statistics computed.
The program is what we call menu-driven. In other words, the user is presented with a menu of
choices (see below) for actions to be taken. This is exactly what most graphical user interface
programs, like Windows applications such as Notepad do. However, this program will be a nongraphical (that is text-based) version similar to what was done before the advent of modern
graphics. As a matter of fact, it would be possible to port the code you will write to a C++
program to add a graphic user interface (GUI) to replace the menu.
Learning Objectives
This project is intended to help you achieve the following learning objectives:
1. How to write modular programs (multiple .c and .h files that are linked together)
2. How to follow a design pattern in planning and implementing the code
3. How to do simple unit testing of individual modules and functions
4. How to incrementally develop and test and then integrate a system
5. What is involved in developing a full application
6. How to work in pairs to learn and produce working applications (later in the series you
will work in teams)
7. (C objectives) Working with pointers, heap memory dynamic allocation, standard library
functions
These are extremely important objectives and what you learn here will be used in many of your
future courses.
Product Specification
I am acting as the chief product designer (in future courses you will learn more about designing
complex systems). I have developed a functional specification for this product that you will use
as the basis for developing the system. This is a fairly simple and straightforward specification
but you will find the implementation and testing will be a challenge for your current level of skill
at programming.
The product to be produced shall conform to the following.
Overview
The product will have the following general functionality.
StatPac – Statistical Package
Routines for capturing and storing data, computing statistics, and generating
reports
Program
Main Menu Sub-menus
File
Create new data file
Open an existing file
Rename file
Copy file to…
Print file contents
Delete file
Edit
Add data
Edit item
Delete item
Search
Search for item
Search and edit
Sort data
Compute
Update all
Compute mean
Compute median
Compute variance
Compute standard deviation
Report
Display statistics
Print statistics
Files and Functions
Below is the list of .c files that will be created along with the major functions that will be built
into those files. The first three items in the list have “man” (manual) pages that provide details as
required on each object. Use these files as examples of how to produce external documents for
the application. There will be a man page for each major file and function.
statpac.c
main() prints menu, gets user selections, calls appropriate submenu
functions.
main_menu()
file.c
file_menu() prints submenu, gets user selections, calls appropriate
functions
file_create() gets user input for new file name, checks to make sure
no file of that type already
exists, if it doesn’t exist it creates the file and initializes the
header lines (see below)
file_open() opens an existing file if it can
file_rename() allows the user to change the name of an existing file
to an alternate
file_copy() makes a copy of an existing file. The new file must have a
different legal filename
file_print() copies a data file to a formatted version suitable for
printing on a printer
file_delete() allows the user to delete an existing file; gives the
user a second chance to abort
edit.c
edit_menu() prints the submenu for the edit functions
edit_add() allows a user to add addition data records to an existing
file
edit_edit() allows a user to edit an existing data record
edit_delete() allows a user to delete an existing data record
search.c
search_menu() prints the submenu for searching a data file
search_item() allows the user to find a specific data record if it
exists
search_sort() allows users to sort the data file by data values, entry
or edit dates
compute.c
compute_menu() prints a menu of computing options
compute_update() allows the user to update the computations of all
statistics
compute_mean() computes the mean value only – displays it and inserts
it into the header record
compute_median() same for the median – if the file is not sorted, will
sort an internal array of data
compute_variance() same for variance if mean is updated
compute_stdev() same for std. dev if mean and variance are updated.
report.c
report_menu() prints report submenu
report_display() shows statistics report on the screen
report_print() prints the report to a file for printing
Data File Format
The data file has a particular structure that contains meta-data (data about data) to support file
and data management. The file consists of a “header record”, the first line of a text file, and some
number of data records. The records are stored on disk in what is known as comma separated
format (CSF). The two kinds of records are shown here with respect to the semantics of their
contents and their definitions in the main header file,
statpac.h.
Header Record
Filename
Date of creation
Date of last modification
Owner name
Number of records
State of order (new, entry, sorted)
Mean
Median
Variance
Std. Deviation
Data Record
Date of entry
Date of modification
Data
/* global.h – partial view */
typedef enum{new, entry, sorted} State ; // defines a State type for use
below
typedef struct file_header {
char * filename;
char creation_date[25];
char modification_date[25];
char * owner;
long num_records;
State state;
double mean;
double median;
double variance;
double std_dev;
} FileHeader;
typedef FileHeader *FileHeaderPtr; // defines a pointer type of dynamic
allocation
typedef struct data_record {
char entry_date[25];
char mod_date[25];
double data;
} DataRecord;
typedef DataRecord *DataRecordPtr; // defines a pointer type of dynamic
allocation
File example: dataFile.dat
dataFile.dat,Sat Jan 10 11:15:06 2015,Sun Jan 11 11:15:06
2015,mmuppa,75,entry,0.0,0.0,0.0,0.0
Sun Jan 11 11:15:06 2015,Sun Jan 11 11:15:06 2015,345.67
Sun Jan 11 11:15:06 2015,Sun Jan 11 11:15:06 2015,222.22
\
/
\
/
Sun Jan 11 11:15:06 2015,Sun Jan 11 11:15:06 2015,541.01
The top line (wrapped) is the header record containing all of the meta-data and last computed
statistics. The next 75 lines contain the actual data along with the entry and edit dates. All of the
elements (called fields) are separated by commas. When reading the data from a file you need to
parse the line, extracting each field into its own data variable in one of the above structures. The
last line is actually not a printed line, it is the end-of-file marker, appended by the operating
system when the file is closed after writing.
Project Organization and Requirements
All students will be paired (one team of two) for this project. The project will be done in three
stages. The first stage will involve developing the main() function and the main_menu()
function.
Phase I
Set up the files (.h and .c) indicated above. Develop the main() function in statpac.c and the
main_menu() function. In each of the menu-related files, start with a stub function for that
submenu. E.g. in file.c build a stub called file_menu() that simply returns.
The objective of phase 1 is to complete the implementation of all of the file management
functions.
statpac.c
main() prints menu, gets user selections, calls appropriate submenu
functions.
main_menu()
file.c
file_menu() prints submenu, gets user selections, calls appropriate
functions
file_create() gets user input for new file name, checks to make sure
no file of that type already
exists, if it doesn’t exist it creates the file and initializes the
header lines (see below)
file_open() opens an existing file if it can
file_rename() allows the user to change the name of an existing file
to an alternate
file_copy() makes a copy of an existing file. The new file must have a
different legal filename
file_print() copies a data file to a formatted version suitable for
printing on a printer
file_delete() allows the user to delete an existing file; gives the
user a second chance to abort
Submit your Phase I deliverables as a zip file on Canvas. Each file must have
your names and the appropriate documentation in the header and for each
function. Use the starter code provided for the header files. You must comply
with the naming conventions provided.
