On this page:
+  -ρ
--ρ
*-ρ
/  -ρ
expt-ρ
exp-ρ
log-ρ
abs-ρ
sqrt-ρ
sqr-ρ
sum-ρ
sum-cols-ρ
*-2-1-ρ
argmax-ρ
max-ρ
concat-ρ
concat-n-ρ
dot-product-ρ
dot-product-2-1-ρ
correlate-ρ
8.12

8 Non-differentiable extended numerical functions🔗ℹ

Non-differentiable versions of the functions in Differentiable extended numerical functions. These functions don’t accept duals or return duals. They are primarily used to seek better memory performance where automatic differentiation is not required.

procedure

(+-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
Adds t0 and t1 based on Binary function extension rules.

procedure

(--ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
Subtracts t0 and t1 based on Binary function extension rules.

procedure

(*-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
Multiplies t0 and t1 based on Binary function extension rules.

procedure

(/-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
Divides t0 into t1 based on Binary function extension rules.

procedure

(expt-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
Computes t0 raised to the power of t1 based on Binary function extension rules.

procedure

(exp-ρ t)  tensor?

  t : tensor?
Computes e raised to the power of t based on Unary function extension rules.

procedure

(log-ρ t)  tensor?

  t : tensor?
Computes logarithm of t based on Unary function extension rules.

procedure

(abs-ρ t)  tensor?

  t : tensor?
Computes absolute value of t based on Unary function extension rules.

procedure

(sqrt-ρ t)  tensor?

  t : tensor?
Computes the square root of t based on Unary function extension rules.

procedure

(sqr-ρ t)  tensor?

  t : tensor?
Computes the square of t based on Unary function extension rules.

procedure

(sum-ρ t)  tensor?

  t : tensor?
If t is a tensor of rank 1, return the sum of all the scalar? elements in t.

When t is of rank higher than 1, the Unary function extension rules apply.

procedure

(sum-cols-ρ t)  tensor?

  t : tensor?
If t is a tensor of rank 2, return the sum of all the scalar? elements in t, calculated using the extended addition function +-ρ.

When t is of rank higher than 2, the Unary function extension rules apply.

When t is of rank 1, this function is the same as sum.

The function is undefined if t is a scalar.

procedure

(*-2-1-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
When t0 is a tensor of shape (list m n) and t1 is a tensor of shape (list n), returns a tensor r of shape (list m n) where the m elements of r are formed by multiplying each element of t0 with t1 using the extended multiplication function *-ρ.

When the ranks of t0 and t1 are higher than 2 and 1 respectively, the Binary function extension rules apply.

The function is undefined if t0 has rank less than 2 and t1 has rank less than 1.

procedure

(argmax-ρ t)  tensor?

  t : tensor?
When t is a tensor of rank 1, returns the index of the highest element in t.

When t is of rank higher than 1, the Unary function extension rules apply.

The function is undefined when t is of rank less than 1.

procedure

(max-ρ t)  tensor?

  t : tensor?
When t is a tensor of rank 1, returns the highest element in t.

When t is of rank higher than 1, the Unary function extension rules apply.

The function is undefined when t is of rank less than 1.

procedure

(concat-ρ t u)  tensor?

  t : tensor?
  u : tensor?
When t and u are tensors of rank 1, returns a tensor that is the concatenation of t and u.

Otherwise, the Binary function extension rules apply.

The function is undefined when t or u is of rank less than 1.

procedure

(concat-n-ρ n)  (-> ([t tensor?] [u tensor?]) tensor?)

  n : positive-integer?
When t and u are tensors of rank n, returns a tensor that is the concatenation of t and u.

Otherwise, the Binary function extension rules apply.

The function is undefined when t or u is of rank less than n.

procedure

(dot-product-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
When t0 and t1 are tensors of rank 1 and the same shape, returns a scalar s formed by multiplying each element of t0 with the corresponding element in t1 and summing the results.

When the ranks of t0 and t1 are higher than 1 and 1 respectively, the Binary function extension rules apply.

The function is undefined if t0 has rank less than 1 and t1 has rank less than 1.

procedure

(dot-product-2-1-ρ t0 t1)  tensor?

  t0 : tensor?
  t1 : tensor?
When t0 is a tensor of shape (list m n) and t1 is a tensor of shape (list n), returns a tensor r of shape (list m) where the m elements of r are formed by the dot-product of each element of t0 with t1.

When the ranks of t0 and t1 are higher than 2 and 1 respectively, the Binary function extension rules apply.

The function is undefined if t0 has rank less than 2 and t1 has rank less than 1.

procedure

(correlate-ρ filter-bank signal)  tensor?

  filter-bank : tensor?
  signal : tensor?
When filter-bank is a tensor of shape (list b m d) and signal is a tensor of shape (list n d), returns a tensor r of shape (list n b) which is the correlation of the filter-bank with the signal.

When the ranks of filter-bank and signal are higher than 3 and 2 respectively, the Binary function extension rules apply.