Status Updates From High Performance Python: Pr...

Lists are dynamic arrays; they are mutable and allow for resizing (changing the number of elements that are held).
2. Tuples are static arrays; they are immutable, and the data within them cannot be changed once they have been created.
3. Tuples are cached by the Python runtime, which means that we don’t need to talk to the kernel to reserve memory every time we want to use one.

The chapter on JIT/AOT compilers is awesome! Did not know how many JIT/AOT compilere there are (rather were in 2014/2015)
Didn't know how easy one can compile python to C with cython. Also found more details about numpy's contiguous arrays and how they affect performance. cffi is also pretty cool. I need more details about inlining c functions in python, there was something I didn't get there.

$ pip install memory_profiler
$ ipython
>>> %load_ext memory_profiler
>>> %memit [0] * 10 ** 8
# will tell you how much memory was used up on that line (700+ MB!)

Also guppy.heapy analizes heap allocation per object type!

The chapter on memory profiling is amazing!

More details on dicts: they use a "perturbation index" on collisions (python determines new indices for the given key - open addressing)

$ pip install memory_profiler
$ ipython
In[1]: %load_ext memory_profiler
In[2]: %memit [0] * 10 ** 8

!awesome little tool!

Also, the timeit module disables the garbage collector when running! Very interesting.