Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Use scipy.ndimage.rotate to rotate an image array by an angle in degrees:
from scipy import ndimage
rotated = ndimage.rotate(image, angle=45, reshape=True)The function rotates in the plane defined by two array axes, using spline interpolation. The main choices are whether to expand the output to fit the rotated image, how to interpolate pixel values, and how to handle areas beyond the input’s edges.
Basic image rotation
For an ordinary two-dimensional image, the default axes select its two dimensions. SciPy’s reference demonstrates the call with its sample image:
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsfrom scipy import ndimage, datasets
img = datasets.ascent()
img_45 = ndimage.rotate(img, 45, reshape=False)
full_img_45 = ndimage.rotate(img, 45, reshape=True)In the SciPy v1.18.0 documentation example, img.shape is (512, 512); the fixed-size result is (512, 512), while the expanded result is (724, 724). These are documented example values, not a performance or image-quality benchmark. See the SciPy v1.18.0 rotate API reference.
#1 Best Overall
Choose whether the output can grow
Keep the original dimensions with reshape=False
This keeps the output shape the same as the input. If the rotated image extends beyond those fixed bounds, the corners or other content outside the output area are cropped.
Fit the rotated image with reshape=True
This adjusts the output dimensions to contain the rotated input. The result may be larger than the original, with additional border area around the image. How that area looks depends on the boundary mode and, for constant mode, the fill value.
Select the rotation plane for multidimensional arrays
The axes argument names the two dimensions that define the rotation plane. Its default is (1, 0), which is appropriate for a conventional two-dimensional image. For a multichannel or higher-dimensional array, specify the axes explicitly when the defaults would rotate the wrong dimensions. The function rotates within the selected plane; it does not choose a plane based on image semantics.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Set interpolation and output dtype deliberately
order sets the spline interpolation order and accepts values from 0 through 5. The default is 3, cubic spline interpolation. No one order is best for every image or data type: choose based on the data and the result you need.
By default, SciPy creates an output array with the same dtype as the input. You can pass output as an array to receive the result or as a dtype to request an output type. This matters when the interpolated values should be represented in a particular type.
Prefiltering for orders above 1
With order > 1, the default prefilter=True creates a temporary float64 filtered array. Setting prefilter=False on input that has not already been spline-filtered can make the result slightly blurred. If the input is already spline-filtered, disabling prefiltering avoids filtering it again.
Control what appears beyond the image edges
The default is mode='constant' with cval=0.0. Outside samples use that constant value, and interpolation is not performed beyond the input edge. This can produce black borders for typical image data, but that fill is not necessarily suitable for every image or array. Choose the mode and fill value according to what the array represents.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware match| Mode | Boundary behavior |
|---|---|
reflect |
Reflects about the edge of the last pixel. grid-mirror is a synonym. |
constant |
Uses cval beyond the edge and does not interpolate beyond the input extent. |
grid-constant |
Uses the constant extension and interpolates outside the input extent. |
nearest |
Repeats the last pixel. |
mirror |
Reflects about the center of the last pixel. |
grid-wrap |
Wraps to the opposite edge. |
wrap |
Wraps with overlapping endpoints, so the selected sample at the overlap is ambiguous. |
For constant mode, set a different cval if zero is not an appropriate outside value. For masks, label arrays, or continuous measurements, consider how interpolation and boundary extension affect the meaning of the values rather than treating the array as ordinary color imagery.
Complete API signature
scipy.ndimage.rotate(
input,
angle,
axes=(1, 0),
reshape=True,
output=None,
order=3,
mode='constant',
cval=0.0,
prefilter=True,
)input: array-like data to rotate.angle: angle in degrees.axes: the two integer axes defining the rotation plane.reshape: whether output dimensions adapt to contain the input.output: an output array or dtype; by default the output has the input dtype.order: spline interpolation order from 0 through 5.mode: how samples beyond the input boundary are handled.cval: constant fill value when using constant mode.prefilter: whether to apply the spline prefilter.
The signature and behavior above are from the SciPy v1.18.0 API documentation; check the reference for the version you use because API details can change.
Complex arrays and alternative backends
SciPy’s documentation says complex-valued input is handled by rotating the real and imaginary components independently. It also describes experimental Python Array API Standard support. For SciPy v1.18.0, the reference lists NumPy on CPU, CuPy on GPU, PyTorch on CPU, JAX on CPU without JIT, and Dask on CPU, where the graph is computed. The listed combinations are version-sensitive; consult the versioned API reference rather than assuming all array libraries or devices are supported.
When another ndimage function may fit better
rotate is the direct choice for a fixed-angle rotation in one plane. If the task requires a custom output-to-input coordinate mapping or a broader affine operation, SciPy’s ndimage reference index also lists geometric_transform, map_coordinates, and affine_transform. Choose among them based on whether the operation is a planar rotation or a more general mapping; the index identifies these related tools but does not prescribe a particular one for every task.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
