General guidelines

• comment your code (everybody must understand what is being done)
• NEVER more than one blank line please!
• Align operators vertically to emphasize local program structure and semantics when possible
• Do not use excessive blank spaces (especially in equations)

C/C++

• if/for should follow this:
  • no space fetween if/for and its statement
  • If an if/for holds on one line, then do not use brackets
  • Otherwise, use brackets

    for(int i=0<i<n;i++)  A[i]=i;
    
    for(int i=0<i<n;i++){
       A[i]=B[i];
       B[i]=0;
    }
    
    if(a==0) bool=true;
    
    if(a==0)
       bool=true;
    else if(a==1)
       bool=flase;
    else
       _error_("a=%g not supported",a);
    
    if(a==0){
        output=true;
        c=b
    }
    else{
        output=false
        c=a;
    }
    

• Comments should follow the code indentation and there should not be any blank line between a comment and the code it is referring to

     /*Assigning values of A*/
     for(int i=0<i<n;i++){
  
        /*The comment here is indented*/
        A[i]=i;
     }
  

• Function declaration should hold on one line only

  bool Test(int a,double b,char* c){
  

MATLAB

All MATLAB routines should start with a help (Example and See Also are not mandatory):

Error: Failed to load processor matlab

No macro or processor named 'matlab' found

At the very least, the first line and the Usage should be provided. Use indentations of 3 and 6 spaces. Example:

Error: Failed to load processor matlab

No macro or processor named 'matlab' found

Python

​PeP8 compliance should be used throughout the code with the exceptions below (with flake8 codes):

• We allow lines of any length (E501)
• We don't enforce the space after "#" for comments (E262, E265)
• We still allow form module import * (F403) but we should avoid those if possible
• We bypass warning on undefined function (F405) this can probably be removed if we take care of the one above

flake8 allows you to track and highlight the syntax errors (through Elpy in Emacs, I guess it can be introduced in vim).

If you install flake8, you can also run it in standalone to check the files in a directory:

flake8  --ignore=E262,E265,F403,F405,E405,E501

Docstrings

Adhering to ​PEP8 Docstring Conventions while attempting to mirror the conventions we follow under MATLAB, modules, functions, classes, and method definitions should be documented according to the following protocol:

def FunctionName(inputs):
    """FUNCTIONNAME - one line description

    Extensive description of what is being done, inputs
    outputs, etc...

    Usage:
        outputs = FunctionName(inputs)

    Example:
        md.test = FunctionName(1)

    See Also:
        FunctionName2, FunctionName3, ...
    """

As with MATLAB, at the very least, the first line and the 'Usage' should be provided. Unlike MATLAB, use indentations of 4 and 8 spaces.

MATLAB Built-In Equivalents

The following is a list of MATLAB built-in equivalents in Python:

MATLAB	Python	Notes
s=num2str(A)	s=str(A)	convert numbers to character array

NumPy/SciPy

NumPy and SciPy are used extensively in the Python interface to ISSM to replicate MATLAB-native functionality. When translating modules or tests, for example, from MATLAB to Python, the following sources may come in handy:

• ​NumPy for Matlab users | NumPy
• ​NumPy for MATLAB users | mathesaurus.sourceforge.net

Some notable omissions in the above sources are as follows:

MATLAB	NumPy	Notes
<<cell_array>>{:}	<<np.ndarray>>.flatten()	Flatten a MATLAB cell array or NumPy ndarray.
find(a>0.5)	np.where(a>0.5)[0] np.where(a>0.5,a,a)[0]	Find the indices where (a > 0.5). When only the condition parameters is provided, this function is a shorthand for np.asarray(condition).nonzero(). Further, when working on a 1D array, this array must be passed in as a value for parameters x and y if the desired output is a single array of indices rather than a tuple of arrays. Note as well that a must be of type np.array (or one of its subclasses): a list will not automatically be cast. See also: ​https://numpy.org/doc/stable/reference/generated/numpy.where.html
find('cond1'&'cond2')	np.where(np.logical_and.reduce(('cond1','cond2'))[0]	Find the indices where 'cond1' and 'cond2' are met. The same protocol can be followed for MATLAB's | by instead using logical_or. More than two conditions may be compounded.
B=sortrows(A,column)	B = A[A[:,column].argsort()]	sort rows of matrix or table (MATLAB), or 2D array (NumPy) in ascending order based on the elements in 'column'

Variable/Enum/Function Names

• variables should not use capital letters. Use underscores to make variables more understandable.

• Function names and enums should not use any underscore. Use capital letters to make names more understandable.

Example:

  Input* vx_input=GetInput(inputs,VxInput);