# regress.py - gets x,y data from file named
#    by user, and plots points and best-fit line
# updated by cmc, 1/29/2015

# getData - asks user for filename, then
#    reads x,y data from the file
# returns list of x,y tuples
def getData():
    filename = input("enter input file name --> ")
    dataFile = open(filename, "r")
    data = []
    for line in dataFile:
        xyStrings = line.split()
        x = float(xyStrings[0])
        y = float(xyStrings[1])
        data.append( (x,y) )
    dataFile.close()
    return data

# plotRegression - Miller/Ranum Programming
#    Exercise 4.2 plots points and best-fit line
# assumes data is a list of x,y tuples
def plotRegression(data):
    n = len(data)
    
    # first find means of x and y values
    # key sums, and extremes
    sumX = sumY = sumX2 = sumXY = 0.0
    maxX = minX = data[0][0]
    maxY = minY = data[0][1]
    for x,y in data:
        sumX = sumX + x
        sumY = sumY + y
        sumX2 = sumX2 + x*x
        sumXY = sumXY + x*y
        if x > maxX:
            maxX = x
        if x < minX:
            minX = x
        if y > maxY:
            maxY = y
        if y < minY:
            minY = y
    meanX = sumX / n
    meanY = sumY / n

    # now find slope of line
    numerator = sumXY - n * meanX * meanY
    denominator = sumX2 - n * meanX**2
    slope = numerator / denominator

    # create a turtle and scale the window
    import turtle
    turtle = turtle.Turtle()
    turtle.hideturtle()
    turtle.speed(10)
    screen = turtle.getscreen()
    screen.setworldcoordinates(minX, minY, maxX, maxY)
    
    # plot the points
    for x,y in data:
        turtle.up()
        turtle.goto(x,y)
        turtle.down()
        turtle.dot()

    # plot the best-fit line
    turtle.color("red")
    turtle.up()
    turtle.goto(minX, meanY + slope * (minX - meanX))
    turtle.down()
    turtle.goto(maxX, meanY + slope * (maxX - meanX))
    
    screen.exitonclick()

# main execution - just one statement necessary
plotRegression( getData() )