Pandas: How to add column to multiindexed dataframe?

Question

I was following a brief tutorial on LinkedIn regarding multiindexed pandas dataframes where I was unable to reproduce a seemingly very basic operation (at 3:00). You DO NOT have to watch the video to grasp the problem.

The following snippet that uses a dataset from seaborn will show that I'm unable to add a column to a multiindexed pandas dataframe using the technique shown in the video, and also described in an SO post here.

Here we go:

import pandas as pd
import seaborn as sns

flights = sns.load_dataset('flights')
flights.head()
flights_indexed = flights.set_index(['year', 'month'])

flights_unstack = flights_indexed.unstack()
print(flights_unstack)

Output:

      passengers                                                               
month    January February March April  May June July August September October   November December
year                                                                            
1949         112      118   132   129  121  135  148    148       136     119        104      118
1950         115      126   141   135  125  149  170    170       158     133        114      140  
1951         145      150   178   163  172  178  199    199       184     162        146      166
1952         171      180   193   181  183  218  230    242       209     191        172      194
1953         196      196   236   235  229  243  264    272       237     211        180      201 
1954         204      188   235   227  234  264  302    293       259     229        203      229 
1955         242      233   267   269  270  315  364    347       312     274        237      278  
1956         284      277   317   313  318  374  413    405       355     306        305      336 
1957         315      301   356   348  355  422  465    467       404     347        310      337   
1958         340      318   362   348  363  435  491    505       404     359        362      405
1959         360      342   406   396  420  472  548    559       463     407        362      405
1960         417      391   419   461  472  535  622    606       508     461        390      432

And now I'd like to append a column that shows the sum per month for each year using

flights_unstack.sum(axis = 1)

Output:

year
1949    1520
1950    1676
1951    2042
1952    2364
1953    2700
1954    2867
1955    3408
1956    3939
1957    4421
1958    4572
1959    5140
1960    5714

The two sources mentioned above demonstrate this by using something as simple as:

flights_unstack['passengers', 'total'] = flights_unstack.sum(axis = 1)

Here, 'total' should appear as a new column under the existing indexes. But I'm getting this:

TypeError: cannot insert an item into a CategoricalIndex that is not already an existing category

I'm using Python 3, and so is the author in the video from 2015.

What's going on here?

I've made a bunch of attempts using only values from series above, as well as reshaping, transposing, merging and joining the data bot as pd.Series and pd.DataFrame. And resetting the indexes. I may have overlooked some important detail, and now I'm hoping for a suggestion from some of you.

EDIT 1 - Here's an attempt after the first suggestion from jezrael:

import pandas as pd
import seaborn as sns

flights = sns.load_dataset('flights')
flights.head()

flights_indexed = flights.set_index(['year', 'month'])

flights_unstack = flights_indexed['passengers'].unstack()
flights_unstack['total'] = flights_unstack.sum(axis = 1)

Output:

TypeError: cannot insert an item into a CategoricalIndex that is not already an existing category

Answer 1

Change:

flights_unstack = flights_indexed.unstack()

to:

flights_unstack = flights_indexed['passengers'].unstack()

for remove Multiindex in columns.

---

And last is necessary add_categories by new column name:

flights_unstack.columns = flights_unstack.columns.add_categories(['total'])

flights_unstack['total'] = flights_unstack.sum(axis = 1)
print (df)
       January  February  March  April  May  June  July  August  September  \
month                                                                        
1949       112       118    132    129  121   135   148     148        136   
1950       115       126    141    135  125   149   170     170        158   
1951       145       150    178    163  172   178   199     199        184   
1952       171       180    193    181  183   218   230     242        209   
1953       196       196    236    235  229   243   264     272        237   
1954       204       188    235    227  234   264   302     293        259   
1955       242       233    267    269  270   315   364     347        312   
1956       284       277    317    313  318   374   413     405        355   
1957       315       301    356    348  355   422   465     467        404   
1958       340       318    362    348  363   435   491     505        404   
1959       360       342    406    396  420   472   548     559        463   
1960       417       391    419    461  472   535   622     606        508   

       October  November  December  total  
month                                      
1949       119       104       118   1520  
1950       133       114       140   1676  
1951       162       146       166   2042  
1952       191       172       194   2364  
1953       211       180       201   2700  
1954       229       203       229   2867  
1955       274       237       278   3408  
1956       306       305       336   4003  
1957       347       310       337   4427  
1958       359       362       405   4692  
1959       407       362       405   5140  
1960       461       390       432   5714  

Setup:

import pandas as pd

temp=u"""month;January;February;March;April;May;June;July;August;September;October;November;December

1949;112;118;132;129;121;135;148;148;136;119;104;118
1950;115;126;141;135;125;149;170;170;158;133;114;140
1951;145;150;178;163;172;178;199;199;184;162;146;166
1952;171;180;193;181;183;218;230;242;209;191;172;194
1953;196;196;236;235;229;243;264;272;237;211;180;201
1954;204;188;235;227;234;264;302;293;259;229;203;229
1955;242;233;267;269;270;315;364;347;312;274;237;278
1956;284;277;317;313;318;374;413;405;355;306;305;336
1957;315;301;356;348;355;422;465;467;404;347;310;337
1958;340;318;362;348;363;435;491;505;404;359;362;405
1959;360;342;406;396;420;472;548;559;463;407;362;405
1960;417;391;419;461;472;535;622;606;508;461;390;432"""
#after testing replace 'pd.compat.StringIO(temp)' to 'filename.csv'
df = pd.read_csv(pd.compat.StringIO(temp), sep=";", index_col=[0])
print (df)


df.columns = pd.CategoricalIndex(df.columns)

df.columns = df.columns.add_categories(['total'])
df['total'] = df.sum(axis = 1)

Answer 2

I know this is kind of late but I found the answer to your problem in the FAQs section of the course. Here's what it says:

"Q. What are the issues with Pandas categorical data?

A. Since version 0.6, seaborn.load_dataset converts certain columns to Pandas categorical data (see http://pandas.pydata.org/pandas-docs/stable/categorical.html). This creates a problem in the handling of the "flights" DataFrame used in "Introduction to Pandas/Using multilevel indices". To avoid the problem, you may load the dataset directly with Pandas:

flights = pd.read_csv('https://raw.githubusercontent.com/mwaskom/seaborn-data/master/flights.csv')"

I hope this helps.