I have a pretty big dataframe with column names categories (foreign trade statistics), while the index is a string containing the country code AND the year: w2013 meaning World, year 2013, r2015 meaning Russian Federation, year 2015.
Index([u'w2011', u'c2011', u'g2011', u'i2011', u'r2011', u'w2012', u'c2012',
u'g2012', u'i2012', u'r2012', u'w2013', u'c2013', u'g2013', u'i2013',
u'r2013', u'w2014', u'c2014', u'g2014', u'i2014', u'r2014', u'w2015',
u'c2015', u'g2015', u'i2015', u'r2015'],
dtype='object')
What would be the easiest way to make a multiple index for plotting the various columns - I need a column plotted for each country and each year?
You can try create Multiindex from_tuples - for extract letters use indexing with str.
import pandas as pd
li =[u'w2011', u'c2011', u'g2011', u'i2011', u'r2011', u'w2012', u'c2012',
u'g2012', u'i2012', u'r2012', u'w2013', u'c2013', u'g2013', u'i2013',
u'r2013', u'w2014', u'c2014', u'g2014', u'i2014', u'r2014', u'w2015',
u'c2015', u'g2015', u'i2015', u'r2015']
df = pd.DataFrame(range(25), index = li, columns=['a'])
print df
a
w2011 0
c2011 1
g2011 2
i2011 3
r2011 4
w2012 5
c2012 6
g2012 7
i2012 8
r2012 9
w2013 10
c2013 11
g2013 12
i2013 13
r2013 14
w2014 15
c2014 16
g2014 17
i2014 18
r2014 19
w2015 20
c2015 21
g2015 22
i2015 23
r2015 24
print df.index.str[0]
Index([u'w', u'c', u'g', u'i', u'r', u'w', u'c', u'g', u'i', u'r', u'w', u'c',
u'g', u'i', u'r', u'w', u'c', u'g', u'i', u'r', u'w', u'c', u'g', u'i',
u'r'],
dtype='object')
print df.index.str[1:]
Index([u'2011', u'2011', u'2011', u'2011', u'2011', u'2012', u'2012', u'2012',
u'2012', u'2012', u'2013', u'2013', u'2013', u'2013', u'2013', u'2014',
u'2014', u'2014', u'2014', u'2014', u'2015', u'2015', u'2015', u'2015',
u'2015'],
dtype='object')
df.index = pd.MultiIndex.from_tuples(zip(df.index.str[0], df.index.str[1:]))
print df
a
w 2011 0
c 2011 1
g 2011 2
i 2011 3
r 2011 4
w 2012 5
c 2012 6
g 2012 7
i 2012 8
r 2012 9
w 2013 10
c 2013 11
g 2013 12
i 2013 13
r 2013 14
w 2014 15
c 2014 16
g 2014 17
i 2014 18
r 2014 19
w 2015 20
c 2015 21
g 2015 22
i 2015 23
r 2015 24
If you need convert years to int, use astype:
df.index = pd.MultiIndex.from_tuples(zip(df.index.str[0], df.index.str[1:].astype(int)))
print df.index
MultiIndex(levels=[[u'c', u'g', u'i', u'r', u'w'], [2011, 2012, 2013, 2014, 2015]],
labels=[[4, 0, 1, 2, 3, 4, 0, 1, 2, 3, 4, 0, 1, 2, 3, 4, 0, 1, 2, 3, 4, 0, 1, 2, 3], [0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4]])
Collected from the Internet
Please contact [email protected] to delete if infringement.
Comments