My ultimate goal is to compute the pct_change
for two rows within a (2, 2) multi-index dataframe but I want the pct_change
row to appear below the other two rows. I know about pct_change()
method but that doesn't make a new row so instead I'm computing the pct_change as a separate dataframe and appending the pct_change
row to an existing dataframe. Here's an example dataframe.
df = pd.DataFrame(data={
'A': [94128, 28198, -70.04], 'B': [3627, 1483, -59.11]},
index=pd.MultiIndex.from_tuples([('Label', '(Jun 1, 2014-Mar 31, 2015)'),
('Label', '(Jun 1, 2015-Mar 31, 2016)'),
('Label', '(Jun 1, 2015-Mar 31, 2016)')],
names=['Text', 'Period']))
Looks like this:
A B
Text Period
Label (Jun 1, 2014-Mar 31, 2015) 94128.00 3627.00
(Jun 1, 2015-Mar 31, 2016) 28198.00 1483.00
(Jun 1, 2015-Mar 31, 2016) -70.04 -59.11
This is a multi-index and I want to rename the last Period
row value to say % Change
.
Desired Output:
A B
Text Period
Label (Jun 1, 2014-Mar 31, 2015) 94128.00 3627.00
(Jun 1, 2015-Mar 31, 2016) 28198.00 1483.00
% Change -70.04 -59.11
Is what I'm trying to do even possible given its multi-index complexity?
>>> df.index
MultiIndex(levels=[['Label'], ['(Jun 1, 2014-Mar 31, 2015)', '(Jun 1, 2015-Mar 31, 2016)']],
labels=[[0, 0, 0], [0, 1, 1]],
names=['Text', 'Period'])
I think you can first get_level_values
of level Period
tolist
, remove last value by indexing [:-1]
and add new list ['% Change']
. Last create new MultiIndex
from_tuples
:
print df.index.get_level_values('Period')[:-1].tolist() + ['% Change']
['(Jun 1, 2014-Mar 31, 2015)', '(Jun 1, 2015-Mar 31, 2016)', '% Change']
#change multiindex
new_index = zip(df.index.get_level_values('Text'),
df.index.get_level_values('Period')[:-1].tolist() + ['% Change'])
df.index = pd.MultiIndex.from_tuples(new_index, names = df.index.names)
print df
A B
Text Period
Label (Jun 1, 2014-Mar 31, 2015) 94128.00 3627.00
(Jun 1, 2015-Mar 31, 2016) 28198.00 1483.00
% Change -70.04 -59.11
print df.index
MultiIndex(levels=[[u'Label'],
[u'% Change', u'(Jun 1, 2014-Mar 31, 2015)', u'(Jun 1, 2015-Mar 31, 2016)']],
labels=[[0, 0, 0], [1, 2, 0]],
names=[u'Text', u'Period'])
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.