Pandas restrucutring to longform

Question

I have a dataframe in this format:

{'April': {0: 3266.0, 3: 3044.0, 6: 3607.0},
 'May': {0: 3767.0, 3: 3708.0, 6: 3709.0},
 'June': {0: 4114.0, 3: 3539.0, 6: 4416.0},
 'July': {0: 4544.0, 3: 5176.0, 6: 5298.0},
 'August': {0: 4912.0, 3: 5424.0, 6: 5217.0},
 'September': {0: 5358.0, 3: 5027.0, 6: 5262.0},
 'October': {0: 5265.0, 3: 5163.0, 6: 5597.0},
 'November': {0: 5167.0, 3: 5621.0, 6: 5457.0},
 'December': {0: 3953.0, 3: 4074.0, 6: 4745.0},
 'January': {0: 4235.0, 3: 5792.0, 6: 5067.0},
 'February': {0: 3506.0, 3: 3861.0, 6: 3704.0},
 'March': {0: 3840.0, 3: 3815.0, 6: 3679.0},
 'year': {0: 2015, 3: 2016, 6: 2017}}

The task is to plot a time-series using the year column and respective values of the month from the dataframe. To do that I have tried converting the dataframe to long-form using pandas.wide_to_long and pandas.pivot but was unsuccessful any help will be much appreciated.

Expected output:

year, month, value
2015, January, 4235.0
2015, February, 3506.0
2015, March, 3840.0
... , ..., ...
2017, November, 5457.0
2017, December, 4745.0

`df.melt(id_vars='year', var_name='month')` – mozway Sep 17 '21 at 20:57 — mozway, Sep 17 '21 at 20:57

score 1 · Accepted Answer · answered Sep 17 '21 at 20:55

You need melt:

df.melt('year', var_name='month')

    year      month   value
0   2015      April  3266.0
1   2016      April  3044.0
2   2017      April  3607.0
3   2015        May  3767.0
4   2016        May  3708.0
5   2017        May  3709.0
6   2015       June  4114.0
...

Pandas restrucutring to longform

1 Answers1