mmcky / nyu-econ-370 Goto Github PK

A student asked an excellent question.

If Tuples are Immutable - why then can they contain Mutable Items like Lists? 

This is nicely answered on this stackexchange post.

http://stackoverflow.com/questions/9755990/why-can-tuples-contain-mutable-items

To install the quantecon package, open a terminal (OS X or Linux) or command line window (Windows) and type:

pip install quantecon

This should then show successful installation. Then you can open python or Jupyter notebooks and start using the package by:

import quantecon as qe

Python has a pretty clever automatic memory manager (depending on who you ask) that takes responsibility for managing python's memory use by allocating and de-allocating memory that is used to store objects. The question was asked today if:

'a' == 'a'

is comparing two instances of the data that makes up the character a in memory or if this operation is comparing the same object (physically in memory) and the variable created points to the same information that makes up the letter a.

As it turns out - this is comparing a variable that points the same information in memory. For some simple immutable values python caches them to reduce expensive operations associated with the underlying memory allocation.

In [1]: id('a')
Out[1]: 140045168580400

In [2]: var = 'a'

In [3]: id(var)
Out[3]: 140045168580400

Therefore in this case python caches the object that makes up the letter 'a' and points variables to it in the memory heap managed by python. For more complex objects this isn't necessarily the case.

In [8]: id(2**64)
Out[8]: 140045087723392

In [9]: var = 2**64

In [11]: id(var)
Out[11]: 140045087720712

In [12]: 2**64
Out[12]: 18446744073709551616

In [13]: var
Out[13]: 18446744073709551616

While the values are the same, these objects do not share the same memory id.

Further Reading

http://foobarnbaz.com/2012/07/08/understanding-python-variables/

Great Question!

Not Part of the Course

A stack overflow is usually used in reference to the C programming language and describes a condition whereby a program overflows the C memory stack. This usually involves a program crash condition as the program is trying to access or use memory it is not permitted to do so.

This is related to the infinite or very deep recursion problem as it is a common cause of the stack overflow condition. Eventually the computer would run out of space on the stack and enter an overflow condition. The stack is a piece of memory that is reserved for the use of local variables and function returns.

Python puts protections in place to prevent this from happening - but you can actually alter that behavior using:

import sys
sys.setrecursionlimit(N)  #Where N is larger than what your system can handle (Depents on Hardware)

But in most cases this is generally a bad idea.

To get the course files you may use the GitHub page and download the raw files to your computer when they get updated.

Not Required - this is for students interested in Git which is not a requirement for this course.

Another way is to use git which is a version control application to retrieve the files. This also has the added benefit of being able to retrieve updates to the course repository easily. Git is installed on Linux (usually by default) and OS X. But it is not installed by default on Windows and would require installation.

git clone https://github.com/mmcky/nyu-econ-370.git

will fetch a copy of the entire repo and save it to your computer in the folder: nyu-econ-370/

To get updates

git pull origin master

will pull the latest changes into your repo.

Note: If you make changes to files within the repo - you may experience merge conflicts if I update older files.

I'm working through the intro to Python 2 notebook. All of a sudden I'm seeing "In[*]" whenever I run a line.
No output is produced either.

When I run the same code in a new notebook, it works. What is happening here?

When you download an ipynb file using the RAW button in GitHub it displays the text int the browser. This is json code and contains all of the jupyter notebook cells. You can copy this text into a file and name it with the extension ipynb and it will contain the notebook when using it with Jupyter.

This is a little clunky so another options is:

1. Launch an instance of the mybinder service
2. Navigate to the notebook of interest
3. Right Click and select Download linked file as

Hi, is there a course staff mailing list where we would send our assignments? If not, which emails ID(s) would you prefer we sent the assignments to?

To download a file from GitHub, such as the assignment PDF.

1. Go to the assignments/ folder
2. Click on the assignment that you want (i.e. assignment-1.pdf)
3. Click on the RAW button.

This will download the PDF to your local computer.

There was a question about Markov Chain Monte Carlo (MCMC) in class so I thought I would post some resource links for those that are interested. This is not part of the course.

Software

1. https://pymc-devs.github.io/pymc/ [Python Package]
2. http://mc-stan.org

Video

1. Scipy 2014 Presentation on PyMC https://www.youtube.com/watch?v=XbxIo7ScVzc

If you are having issues with the following import

from pandas_datareader import wb,data

you can fix this by updating your anaconda distribution to the latest or issuing the following command (on the command line) to make sure the package is installed

conda install pandas-datareader

In Assignment #6 Question 4 we are asked about returns and tasked with showing that:

Mathematically though isn't that asking to prove equality between a vector or scalar on the left and a matrix on the right? Any help would be appreciated...

For those of you using Jupyter notebooks for homework assignments - it is possible to use Jupyter to convert your document to LaTeX and then to PDF but this will require the installation of:

1. LaTeX - TexLive (Windows or Linux) or MacTeX (OS x)
2. Pandoc - Utility for converting between document types.

An easier route is to use your operating systems print feature and export it as a PDF

1. OS X = Select File on the OS menubar (not Jupyter's menu bar) and then Export as PDF ...
2. Windows = Select File on the OS menubar (not Jupyter's menu bar) and the Print .. or use Ctrl + P. Then Change Printer destination to Save as PDF. This worked using Chrome on Windows 7 but other versions may have slightly different wording.

If you have trouble printing it out then please submit the assignment via email to me before the class.

Note: If you want to setup LaTex and Pandoc - I am happy to assist during the student consultation hours next Tuesday.

Writing math in Jupyter notebook markdown cells is fairly straight forward once you know the LaTeX commands. Below are some links to resources that will help you learn LaTeX. Once you have specified the math with LaTex then you may enter it into the Jupyter markdown type cell using $ ... $ for inline math and $$ .... $$ for display type formulae.

Checkout the demo notebooks for some examples. You can see the markdown by double clicking on the markdown cells.

Resources:

1. https://en.wikibooks.org/wiki/LaTeX/Mathematics
2. http://jupyter-notebook.readthedocs.org/en/latest/notebook.html#markdown-cells

One of the great features of Jupyter is the live object inspection to obtain information about functions and methods.

To see what functions / methods are available through the objects interface you can:

import numpy as np
np.<tab> #Use the tab key in place of <tab>

To obtain the function signature or the parameter definitions you can:

import numpy as np
np.array(<shift>+<tab> #Replace <shift>+<tab> with the key combination after the open bracket

To get the full docstring:

import numpy as np
np.array(<shift>+<tab>+<tab> #Replace <shift>+<tab> plust another <tab> with the key combination after the open bracket to display more of the docstring

Another way to access this information is using the ? character in Jupyter

import numpy as np
np.array?

then execute the statement to obtain an information window at the bottom of your browser.