CODEWITHSUNDEEP
pythonpython-3.xpandasmatplotlibjupyter-notebook
ken
ken

Reputation: 21

How do I plot a graph for a specific subset of a dataframe

I've been trying to plot a graph for a specific subset of my data. My data has got the data from the years 1960 to 2018. However, I am only interested in plotting my histogram using only a specific column's variable and the rows that display data from 1981 onwards.

So far I've tried plotting using 2 variables

x = df1y.index

which returns the values:

Int64Index([1981, 1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991,
        1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002,
        2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013,
        2014, 2015, 2016, 2017],
       dtype='int64', name=' Time ')

and 

y = df1.iloc[21:58, [15]] ## 21 to 58 are rows for 1981 to 2017 while column 15 refers to the column I've been trying to get

y returns:

Resident Live-births(Number)
Time    
1981    41,000
1982    41,300
1983    39,300
1984    40,200
1985    41,100
1986    37,159
1987    42,362
1988    51,537
1989    46,361
1990    49,787
1991    47,805
1992    47,907
1993    48,739
1994    48,075
1995    46,916
1996    46,707
1997    45,356
1998    41,636
1999    41,327
2000    44,765
2001    39,281
2002    38,555
2003    35,474
2004    35,135
2005    35,528
2006    36,272
2007    37,074
2008    37,277
2009    36,925
2010    35,129
2011    36,178
2012    38,641
2013    35,681
2014    37,967
2015    37,861
2016    36,875
2017    35,444

After keying in 

x = df1y.index
y = df1.iloc[21:58, [15]]
plt.plot(x, y, 'o-')

I've received an error:

TypeError: unhashable type: 'numpy.ndarray'

Upvotes: 1

Views: 3902

Answers (1)

SpghttCd
SpghttCd

Reputation: 10860

use

y = df1.iloc[21:58, 15].values

to do it the way you tried to walk

However, normally you don't want to calculate the subset indices by yourself, so consider sth like

y = df1.loc[df1.index > 1981, 'name_of_your_column_15_here'].values

to get the numpy array of the (y-)values you want to have.

And to get some more convenience, just try and apply .plot() directly to the series (works also with whole dataframes) and look what happens...

idx_slice = df1.index > 1981
df1.loc[idx_slice, 'name_of_your_column_15_here'].plot()

Upvotes: 1

Related Questions