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	t	19
1	x	23
2	y	24
3	l	11
4	n	13
5	u	20
6	q	16
7	g	6
8	a	0
9	d	3

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

#	ascii	index
0	m	12
1	o	14
2	b	1
3	y	24
4	s	18
5	v	21
6	a	0
7	p	15
8	w	22
9	j	9

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	t	19	t
1	y	24	y
2	q	16	q
3	a	0	a
4	d	3	d
5	k	10	k
6	m	12	m
7	j	9	j
8	v	21	v