Optimisation Episodes Draft

Still requires internal review and additional exercises/flow changes.

config.yaml

@@ -59,14 +59,21 @@ contact: '[email protected]'
 
 # Order of episodes in your lesson
 episodes: 
-- introduction.md
 - profiling-introduction.md
 - profiling-functions.md
 - profiling-lines.md
 - profiling-conclusion.md
+- optimisation-introduction.md
+- optimisation-use-latest.md
+- optimisation-memory.md
+- optimisation-list-tuple.md
+- optimisation-dict-set.md
+- optimisation-minimise-python.md
+- optimisation-conclusion.md
 
 # Information for Learners
 learners: 
+- setup.md
 
 # Information for Instructors
 instructors: 

episodes/fig/cpython_list_allocations.py

@@ -0,0 +1,29 @@
+"""
+Construct a graph demonstrating the number of re-allocations
+if growing a CPython List with append()
+"""
+t = 0
+a = 0
+x = range(1000000)
+
+allocs = [0 for x in range(1000000)]
+while t < 1000000:
+    t += 1
+    t_old = t
+    # https://github.com/python/cpython/blob/a571a2fd3fdaeafdfd71f3d80ed5a3b22b63d0f7/Objects/listobject.c#L74
+    t = t + (t >> 3) + 6
+    a+=1
+    for k in range(t_old, t+1):
+        if k < len(allocs):
+            allocs[k] = a
+
+import matplotlib.pyplot as plt
+
+plt.plot(x, allocs)
+
+plt.xlabel("Appends")
+plt.ylabel("Resizes")
+#plt.ticklabel_format(style='plain')
+#plt.xscale('symlog')
+plt.savefig('cpython_list_allocations.png')
+plt.show()

episodes/fig/latency.py

@@ -0,0 +1,20 @@
+"""
+Construct a graph demonstrating the difference in latencies relevant to code execution
+Inspired by
+https://gist.github.com/jboner/2841832
+Using data from
+https://www.intel.com/content/www/us/en/developer/articles/technical/memory-performance-in-a-nutshell.html
+"""
+import matplotlib.pyplot as plt
+label = ["L1 cache", "L2 cache", "L3 cache", "RAM", "SSD", "HDD", "Ldn->Ca->Ldn"]
+latency_ns = [1, 4, 40, 80, 8000, 80000, 140000000]
+
+plt.figure().set_figheight(2)
+plt.barh(label, latency_ns)
+
+plt.xlabel("Latency (nanoseconds)")
+#plt.ylabel("")
+plt.xscale('symlog')
+plt.tight_layout()
+plt.savefig('latency.png')
+plt.show()

episodes/optimisation-conclusion.md

@@ -0,0 +1,34 @@
+---
+title: "Optimisation Conclusion"
+teaching: 0
+exercises: 0
+---
+
+:::::::::::::::::::::::::::::::::::::: questions
+
+- What has been learnt about writing performant Python?
+
+::::::::::::::::::::::::::::::::::::::::::::::::
+
+
+::::::::::::::::::::::::::::::::::::: objectives
+
+- Review what has been learnt about writing performant Python
+
+::::::::::::::::::::::::::::::::::::::::::::::::
+
+This concludes the optimisation portion of the course.
+
+An overview of how Python operates and the most important practices for achieving performant code have been introduced.
+
+Hopefully with the information from this course you will be in a better position to investigate and optimise the performance of your own code.
+
+This course's website can be used as a reference manual when profiling your own code.
+
+[Let us know](https://github.com/RSE-Sheffield/pando-python) what you think we've missed, so we can improve it too!
+
+::::::::::::::::::::::::::::::::::::: keypoints
+
+<!-- todo -->
+
+::::::::::::::::::::::::::::::::::::::::::::::::