Frequency Tables and Writing CSVs

What is Frequency?

Frequency: The number of times a value occurs within the data.

candybars_mini

	name	weight	available_canada_america
0	Coffee Crisp	50	Canada
1	Butterfinger	184	America
2	Skor	39	Both
...	...	...	...
4	Twix	58	Both
5	Reeses Peanutbutter Cups	43	Both
6	3 Musketeers	54	America

7 rows × 3 columns

What is a Frequency Table?

available_canada_america
Both       3
Canada     2
America    2
Name: count, dtype: int64

Before we explain what a frequency table is, you must know what frequency means first.

Frequency is simply put, the number of times a value occurs within the data. Let’s look at an example using our candybars dataset.

If we count the number of times the value Both appears in the available_canada_america column, we get 3 times. This is the frequency of the value both.

A frequency table is a manner of displaying all the possible values of a column in our dataframe and the number of occurrences (frequencies) of each value.

For our sample data, a frequency table for the available_canada_america column would look like this:

mfr_column = cereal['mfr']
mfr_column

0     N
1     Q
2     K
     ..
74    R
75    G
76    G
Name: mfr, Length: 77, dtype: object

mfr_freq = mfr_column.value_counts()
mfr_freq

mfr
K    23
G    22
P     9
     ..
Q     8
N     6
A     1
Name: count, Length: 7, dtype: int64

If we want to get a frequency table of a categorical column, there are a few steps that need to be followed.

Up until now, we discussed getting a single column from a dataframe using double square brackets - df[['column name']].

For frequency tables, however, we only use single brackets to obtain the column values.

We saved the object in this example here to an object named mfr_column in the same way that we have done this before.

Now we can use .value_counts() on this mfr_column variable to reference it, and we can obtain the frequency value for the different categories in that variable.

mfr_col_dataframe = cereal[['mfr']]
mfr_col_dataframe

	mfr
0	N
1	Q
2	K
...	...
74	R
75	G
76	G

77 rows × 1 columns

mfr_col_dataframe.value_counts()

mfr
K      23
G      22
P       9
       ..
Q       8
N       6
A       1
Name: count, Length: 7, dtype: int64

value_counts() can also be called on a DataFrame to count unique rows.

Saving a dataframe

mfr_freq.to_csv('data/mfr_frequency.csv', index=False)

Sometimes it is useful to save a new dataframe to a file like a csv file for future use by you or somebody else.

We can do this using a method called .to_csv().

We put our desired csv file name in quotations within the parentheses and follow it with the argument index=False so we don’t export our index column which is just a column of numbers.

Let’s apply what we learned!