Case study–Python for non-graphic EDA
A team member wants to determine whether the viscosity of the resin batches produced by Reactor A differs from that by Reactor B. She collects viscosity data from 39 batches produced by each of the reactors. She first performs non-visual EDA to view the main characteristics of the dataset.
In [1]:
import numpy as np
import pandas as pd
from scipy import stats
In [2]:
df = pd.read_excel('case moods medain.xlsx')
In [3]:
df.head()
Out[3]:
|
Reactor A
|
Reactor B
|
|
0
|
1100
|
660
|
|
1
|
880
|
1430
|
|
2
|
1100
|
1045
|
|
3
|
1210
|
935
|
|
4
|
770
|
1210
|
In [4]
:
df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 39 entries, 0 to 38
Data columns (total 2 columns):
Reactor A 39 non-null int64
Reactor B 39 non-null int64
dtypes: int64(2)
memory usage: 752.0 bytes
In [4]:
df.describe()
Out[4]:
|
Reactor A
|
Reactor B
|
|
count
|
39.000000
|
39.000000
|
|
mean
|
985.769231
|
965.384615
|
|
std
|
196.045170
|
215.640004
|
|
min
|
713.000000
|
603.000000
|
|
25%
|
824.000000
|
878.000000
|
|
50%
|
935.000000
|
935.000000
|
|
75%
|
1102.000000
|
1047.000000
|
|
max
|
1322.000000
|
1432.000000
|
|
|
|
The dataset has two columns, and each of them has 39 data points. The data in both columns are continuous, and all data points are integers. The mean and median values from the two reactors are quite different from each other.