Copy and add the last line of a python pandas data frame on to itself with updated index

Question

I have a dataframe such as:

---

2013-07 114.60 89.62 125.64

2013-08 111.55 88.63 121.57

2013-09 108.31 86.24 117.93

---

index is YY-MM date series I would like to copy and add the last row to the original dataframe with a new updated index. The new dataframe should look like:

---

2013-07 114.60 89.62 125.64

2013-08 111.55 88.63 121.57

2013-09 108.31 86.24 117.93

2013-10 108.31 86.24 117.93

---

how can I do this?

Answer 1

This is how I parsed your data (easy, but you really should have code snippets describing the data in your question):

In [1]: df = pd.read_csv('in.txt', index_col=0, sep=' ', header=None, parse_dates=[0])

In [2]: df
Out[2]: 
                 1      2       3                                                                                                                                                             
0                                                                                                                                                                                             
2013-07-01  114.60  89.62  125.64                                                                                                                                                             
2013-08-01  111.55  88.63  121.57                                                                                                                                                             
2013-09-01  108.31  86.24  117.93

Now, using concat/append and slicing, you can re-add the last row under a new date with:

In [3]: new_date = pd.datetools.to_datetime('2013-10')

In [3]: new_data = pd.DataFrame(df[-1:].values, index=[new_date], columns=df.columns)

In [4]: df = df.append(new_data)

In [5]: df
Out[5]: 
                 1      2       3                                                                                                                                                                                                   
2013-07-01  114.60  89.62  125.64                                                                                                                                                                                                   
2013-08-01  111.55  88.63  121.57                                                                                                                                                                                                   
2013-09-01  108.31  86.24  117.93                                                                                                                                                                                                   
2013-10-01  108.31  86.24  117.93  

Note, however, that adding data row by row is not the recommended way - it is better to do appends on lower-level structures, such as lists and dicts (which are faster at individual appends), and convert the data to a DataFrame at bulk when you actually need to analyse it.