General key differences

Python Conventions

We make extensive use of NumPy throughout the library so please see their documentation for MATLAB users. Otherwise some highlights are:

• 0-indexing in python vs 1-indexing in MATLAB

Copying

Copying a pyttb tensor works differently than MATLAB. For example in MATLAB, copying a tensor Y as Y = X returns a tensor Y that is independent of X. Changing the value of Y does not change the value of X. However, the same syntax in pyttb, Y = X, returns a shallow copy of X; the shallow copy Y is a reference to X. For that reason, each pyttb tensor class provides a copy() method that returns a deep copy Y that is independent of X, which is called as Y = X.copy().

Data members

MATLAB name	pyttb name	Calling convention
size	shape	X.shape

Methods

MATLAB name	pyttb name	Calling convention
and	logical_and	X.logical_and(Y)
disp	__str__	X
display	__repr__	print(X)
eq	__eq__	X == Y
ge	__ge__	X >= Y
gt	__gt__	X > Y
ldivide	__truediv__	X / Y
le	__le__	X <= Y
lt	__lt__	X < Y
minus	__sub__	X - Y
mldivide	__truediv__	X / Y
mrdivide	__rtruediv__	X / Y
mtimes	__rmul__	a * X
ne	__ne__	X != Y
not	logical_not	X.logical_not(Y)
or	logical_or	X.logical_or(Y)
plus	__add__	X + Y
power	__pow__	X ** a
rdivide	__rtruediv__	X / Y
subsasgn	__setitem__	X[index] = a
subsref	__getitem__	X[index]
tenfun	tenfun	X.tenfun(lambda x: x + 1)
times	__mul__	X * Y
uminus	__neg__	-X
uplus	__pos__	+X
xor	logical_xor	X.logical_xor(Y)

MATLAB methods not included in pyttb

• datadisp

• isscalar

• transpose

• tocell