Align datasets

Shows how to combine information from two datasets describing the same cases, but not necessarily in the same order.

# Author: Christian Brodbeck <christianbrodbeck@nyu.edu>
import random
import string

from eelbrain import *


# Generate a dataset with known sequence
ds = Dataset()
ds['ascii'] = Factor(string.ascii_lowercase)
# Add an index variable to the dataset to later identify the cases
ds.index()

# Generate two shuffled copies of the dataset (and print them to confirm that
# they are shuffled)
ds1 = ds[random.sample(range(ds.n_cases), 15)]
ds1.head()

#	ascii	index
0	o	14
1	n	13
2	q	16
3	v	21
4	c	2
5	z	25
6	b	1
7	f	5
8	k	10
9	s	18

ds2 = ds[random.sample(range(ds.n_cases), 16)]
ds2.head()

#	ascii	index
0	m	12
1	b	1
2	t	19
3	e	4
4	h	7
5	f	5
6	p	15
7	j	9
8	v	21
9	s	18

Align the datasets

Use the "index" variable added above to identify cases and align the two datasets

ds1_aligned, ds2_aligned = align(ds1, ds2, 'index')

# show the ascii sequences for the two datasets next to each other to
# demonstrate that they are aligned
ds1_aligned['ascii_ds2'] = ds2_aligned['ascii']
ds1_aligned

#	ascii	index	ascii_ds2
0	o	14	o
1	n	13	n
2	v	21	v
3	b	1	b
4	f	5	f
5	k	10	k
6	s	18	s
7	p	15	p
8	t	19	t
9	h	7	h
10	e	4	e