Free tools Windows power users keep installed
One-click scans. No signup required.
Use scipy.stats.skew to calculate sample skewness for an array, either along an axis or across its flattened values. By default, it returns the biased Fisher–Pearson moment coefficient; set bias=False for the adjusted coefficient. The sign describes tail asymmetry, but skewness alone is not a statistical test.
Calculate skewness in Python
Import skew from scipy.stats and pass it a sequence or array:
from scipy.stats import skew
values = [2, 8, 0, 4, 1, 9, 9, 0]
result = skew(values)
print(result) # 0.2650554122698573This matches an example in the SciPy v1.18.0 skew reference. That page also shows that skew([1, 2, 3, 4, 5]) returns 0.0.
What SciPy’s skewness value means
SciPy defines the default coefficient as g₁ = m₃ / m₂3/2, where m₂ and m₃ are the second and third central moments calculated using denominator N. For a unimodal continuous distribution, a positive value indicates greater weight in the right tail; a value below zero indicates greater weight in the left tail. A normally distributed sample should have skewness about zero, though a finite sample need not produce exactly zero.
#1 Best Overall
Skewness is a descriptive measure, not a test of whether the distribution is statistically skewed. SciPy’s skew reference points to scipy.stats.skewtest for assessing statistically whether skewness is close enough to zero. The SciPy v1.18.0 statistical functions index also lists related distribution checks such as normaltest and jarque_bera.
Choose biased or adjusted skewness
The default bias=True returns the moment coefficient g₁. To apply SciPy’s bias adjustment, use bias=False; the adjusted Fisher–Pearson coefficient is G₁ = √(N(N−1))/(N−2) × m₃/m₂3/2.
Rank #2
- This guide is a perfect overview for the topics covered in introductory statistics courses.
from scipy.stats import skew
biased = skew(values) # bias=True by default
adjusted = skew(values, bias=False)The choice changes the estimator, not the direction convention: positive still means greater right-tail weight under SciPy’s stated interpretation. Use the same setting when comparing results across datasets or analyses. SciPy’s describe reference likewise documents that skewness and kurtosis calculations can be bias-corrected.
Control axes, NaNs, and output shape
For a multidimensional array, axis=0 is the default, so SciPy calculates skewness along axis 0. Use another axis to select a different direction, or set axis=None to flatten the input before calculating one value. keepdims=True leaves reduced axes in the result as dimensions of length one, which can help with broadcasting.
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 matchRank #3
| Option | Behavior | Example |
|---|---|---|
axis=0 |
Calculate along axis 0 (default). | skew(data, axis=0) |
axis=1 or another axis |
Calculate along the selected axis. | skew(data, axis=1) |
axis=None |
Flatten the input and calculate a single result. | skew(data, axis=None) |
nan_policy='propagate' |
Default: an affected axis slice returns NaN. | skew(data, nan_policy='propagate') |
nan_policy='omit' |
Ignore NaNs; a slice with too few usable values returns NaN. | skew(data, nan_policy='omit') |
nan_policy='raise' |
Raise ValueError if NaNs are present. |
skew(data, nan_policy='raise') |
keepdims=True |
Retain reduced axes as dimensions of length one. | skew(data, axis=1, keepdims=True) |
These options are documented in the SciPy v1.18.0 function reference. Select the NaN policy to match your workflow rather than assuming missing values are removed automatically.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Degenerate data and compatibility
If all values in a slice are equal, SciPy’s current reference says the result is NaN. This is a zero-variation case, so do not treat the NaN as a measured skew direction.
The SciPy v1.18.0 reference labels its Array API support experimental. It lists NumPy on CPU, CuPy on GPU, PyTorch on CPU and GPU, JAX on CPU and GPU, and Dask on CPU. These are the combinations listed for that version; compatibility may differ in other SciPy versions or environments. See the versioned compatibility table before relying on a particular backend.
Quick Recap
Best Value
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.
