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)]
print(ds1.head())
ds2 = ds[random.sample(range(ds.n_cases), 16)]
print(ds2.head())

Out:

ascii   index
-------------
d       3
k       10
t       19
h       7
n       13
m       12
a       0
l       11
w       22
f       5
ascii   index
-------------
d       3
g       6
v       21
n       13
z       25
t       19
m       12
o       14
b       1
a       0

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']
print(ds1_aligned)

Out:

ascii   index   ascii_ds2
-------------------------
d       3       d
t       19      t
n       13      n
m       12      m
a       0       a
f       5       f
q       16      q
g       6       g
v       21      v
o       14      o