Intermediate

Aggregations & Statistics

sum, mean, and friends, plus what axis actually means.

axis confuses almost everyone at first. Read it as "the axis that collapses": axis=0 collapses rows (you get one result per column), axis=1 collapses columns (one result per row). No axis at all reduces the whole array to a single number.

Function	Returns	Example
`a.mean()`, `a.std()`, `a.var()`	average, standard deviation, variance	`array([1,2,3]).mean()` → `2.0`
`a.min()`, `a.max()`	smallest / largest value	`array([3,1,2]).max()` → `3`
`a.argmin()`, `a.argmax()`	position of the smallest / largest value	`array([3,1,2]).argmin()` → `1` (index of the `1`)
`np.median(a)`, `np.percentile(a, 75)`	median, 75th percentile	`np.median([1,2,3,4])` → `2.5`
`np.cumsum(a)`	running total	`np.cumsum([1,2,3])` → `array([1, 3, 6])`

💡 Tip: Every aggregation has a NaN-safe twin: nanmean, nansum, nanmax, nanstd, nanmedian, nanquantile. Reach for them whenever missing data is possible.

Now practice it

Exercises that use what this lesson just covered.

Min and MaxEasy Mean ValueEasy Peak-to-Peak Range per RowEasy Correlation Matrix Between RowsMedium Weighted AverageEasy Histogram Bin CountsEasy

← Shape Manipulation & Broadcasting Random Numbers & File I/O →