Creating serial graphs with arcpy
Well, are you struggling slightly(almost there, just something in your code is not right) or are you lost (nothing works like it should) 
Perhaps i can help you if you have something that just isn't quite working, if you are just lost,or not sure how to go about this.
My script uses numpy, matplotlib, and scipy, so you would need to install those modules in python (not sure if numpy is installed with arcpy, but i think it is.)
What my script does:
1. So you have a point feature class with a field that matches the DDP sheet number that it falls in
2. It then selects those features, and creates a graph, for that page generalizing for missing data to get rid of the jagged edges, as a png in the output folder
3. Then it inserts the graph onto the page and exports, and moves on to the next.
So at the end, in the output folder there are the pdf's with the graph inserted on them and all of the png's in case you want them for something else.
I really only use this as a function now to create the graphs, I have other functions in my tools that inserts graphs, graphics, pictures, labels, etc on the map. I tried to comment it to make it easier to understand, but let me know if you have any questions.
import arcpy,os,glob
import matplotlib.pyplot as plt
import numpy as np
from scipy.interpolate import spline
mxd = arcpy.mapping.MapDocument(r"<MXD file with Data Driven Pages>")
outputFolder = "C:/<ouput location>/"
# there should be a picture element on your page at the location where you want the graph inserted
pictureElementName = "<Picture Element Name>"
eleFieldFeature = "<Elevation Point Feature Class>"
eleField = "<Elevation Field>"
# distField can be whatever field you want for the X axis
distField = "<Distance Field>"
df = arcpy.mapping.ListDataFrames(mxd, "Layers")[0]
pict = arcpy.mapping.ListLayoutElements(mxd, "PICTURE_ELEMENT", pictureElementName)[0]
lyr = arcpy.mapping.ListLayers(mxd, eleFieldFeature, df)[0]
for pageNum in range(1, mxd.dataDrivenPages.pageCount + 1):
mxd.dataDrivenPages.currentPageID = pageNum
# change this SeqId to match the field name that you have in the DDP featureclass
# this is redundant, but it works all the same and I didn't feel like changing it. :)
pageName = mxd.dataDrivenPages.pageRow.getValue("SeqId")
# change this SeqId to match the field name that you have in the Elevation featureclass
arcpy.SelectLayerByAttribute_management(lyr, "NEW_SELECTION", ' "SeqId" = '+ repr(pageName))
# create array of selected
arr = arcpy.da.FeatureClassToNumPyArray(lyr, (eleField, distField))
# set min and max for the graph from the array
mineleField = arr[eleField].min()
maxeleField = arr[eleField].max()
minStation = arr[distField].min()
maxStation = arr[distField].max()
# matplotlib stuff
x = []
y = []
elevPNG = outputFolder+"\elev"+repr(pageNum)+".png"
fig = plt.figure(figsize=(32, 2.8))
table = lyr
fields = [distField, eleField]
with arcpy.da.SearchCursor(table, fields) as rows:
for row in rows:
x.append(row[0])
y.append(row[1])
# the next 7 lines generalize, smooth the line in case of missing data
x_sm = np.array(x)
y_sm = np.array(y)
x_smooth = np.linspace(x_sm.min(), x_sm.max(),200)
y_smooth = spline(x,y,x_smooth)
# subtract/add 10 to the min max of the graph so elevation fits, can increase or decrease as you need
plt.ylim((mineleField-10,maxeleField+10))
plt.xlim((minStation,maxStation))
plt.plot(x_smooth,y_smooth, color='red', linewidth=3)
## the next 3 is your X label, Y Label, and Graph Title
plt.xlabel('Distance')
plt.ylabel('Elevation')
plt.title('Landscape Profile')
plt.grid(True)
fig.savefig(elevPNG, bbox_inches='tight', dpi=(100))
pict.sourceImage = elevPNG
# end of matplotlib stuff
print "Exporting page {0} of {1}".format(str(mxd.dataDrivenPages.currentPageID), str(mxd.dataDrivenPages.pageCount))
arcpy.mapping.ExportToPDF(mxd, outputFolder + str(pageName) + "_W_GRAPH.pdf", resolution=150, image_quality="NORMAL")
plt.clf()
plt.close()
del fig
del mxd
