Computing statistics

Compute a statistical summary of the data.

Column statistics

We'll compute a summary for each column of the clean data. The day column (0) is only a sequence number and its statistics are meaningless, so it is skipped with veri[:, 1:]:

olcum = veri[:, 1:]          # temperature and humidity
print(olcum.mean(axis=0))    # [23.   48.75]
print(olcum.max(axis=0))     # [26. 60.]

A per-feature summary lets you see the data's range and catch faulty values (for example 99 °C).

argmax: the hottest day

argmax gives the position of the largest value. With that position, the matching value is read from another column:

i = veri[:, 1].argmax()
print(i)              # 3  (0-based)
print(veri[i, 0])     # 4.0  -> day 4

argmin does the same for the smallest value.

Exercise Compute the summary

Print the mean, standard deviation, minimum and maximum of the temperature column (veri[:, 1]) in a single print. Then find the hottest day with argmax and print it as an integer (int).

Show example solution

import io
import numpy as np

csv = "gun,sicaklik,nem\n1,20,40\n2,22,\n3,,60\n4,26,50\n5,24,45\n"
veri = np.genfromtxt(io.StringIO(csv), delimiter=",", skip_header=1)
veri = np.where(np.isnan(veri), np.nanmean(veri, axis=0), veri)

s = veri[:, 1]
print(s.mean(), s.std(), s.min(), s.max())
print(int(veri[s.argmax(), 0]))

Test yourself

How does filling with the mean affect the column's mean?
- It decreases it
- It makes it nan
- It doesn't change it
- It increases it
Show answer

It doesn't change it — Adding a value equal to the mean keeps the mean; it does shrink the standard deviation slightly.
Find the position of the lowest humidity.
```
i = veri[:, 2].____()
```
Show answer
```
i = veri[:, 2].argmin()
```
Match the expressions.

veri[:, 1:]

Columns except the day
max(axis=0)

Column maxima
argmax()

Position of the maximum
veri[i, 0]

That row's day

Next lesson: Printing results

We Are Enhancing Your Experience

Cookie Preferences

Strictly Necessary Cookies

Analytical Cookies