search
Search
Publish
Guest 0reps
Thanks for the thanks!
close
account_circle
Profile
exit_to_app
Sign out
search
keyboard_voice
close
Searching Tips
Search for a recipe: "Creating a table in MySQL"
Search for an API documentation: "@append"
Search for code: "!dataframe"
Apply a tag filter: "#python"
Useful Shortcuts
/ to open search panel
Esc to close search panel
to navigate between search results
d to clear all current filters
Enter to expand content preview Doc Search Code Search Beta SORRY NOTHING FOUND!
mic
Start speaking... Voice search is only supported in Safari and Chrome.
Shrink
Navigate to
A
A
share
thumb_up_alt
bookmark
arrow_backShare Twitter Facebook

# NumPy | amax method

Programming
chevron_right
Python
chevron_right
NumPy
chevron_right
Documentation
schedule Mar 10, 2022
Last updated
local_offer PythonNumPy
Tags

Numpy's `amax(~)` method returns the largest value in the Numpy array. The maximums can also be computed row-wise and column-wise.

# Parameters

1. `a` | `array_like`

The input array.

2. `axis`link | `None` or `int` | `optional`

The allowed values are as follows:

Parameter value

Meaning

axis=0

Maximum computed column-wise

axis=1

Maximum computed row-wise

None

Maximum computed from entire array

By default, `axis=None`.

3. `initial`link | `int` | `optional`

If the computed maximum is less than `initial`, then `initial` will be returned instead.

4. `where`link | `array-like` of `booleans` | `optional`

Instead of considering all the values, we can choose which values to consider by providing this parameter. Only values corresponding to `True` in the mask will be considered.

# Return value

A scalar is returned if the axis parameter is not supplied. Otherwise, a Numpy array is returned.

# Examples

## Maximum of the entire array

``` np.amax([[2,5],[1,3]]) 5 ```

## Maximum of each column

``` np.amax([[2,5],[1,3]], axis=0) array([2, 5]) ```

## Maximum of each row

``` np.amax([[2,5],[1,3]], axis=1) array([5, 3]) ```

## Handling maximum values

When your array contains missing values (e.g. NaNs), then `NaN` is returned:

``` np.amax([2,np.NaN,1,3]) nan ```

If you want to ignore missing values, then use `np.nanmax(~)` method instead.

## Passing in initial parameter

``` np.amax([[2,5],[1,3]], initial=8) 8 ```

Here, the computed maximum is 5, yet it is smaller than the provided value of initial (i.e. 8), so 8 is returned instead.

## Passing in a boolean mask

Instead of considering all the values, we can choose which values to compute the maximum of by providing a mask:

``` np.amax([2,5,3,4], where=[True,False,False,True], initial=-1) 4 ```

Here, although 5 is technically the largest value, it is ignored since its corresponding value in the mask is `False`. Note that we need to supply the parameter `initial` here, which will be the returned value if the maximum cannot be computed (e.g. when the mask is all `False`).