There is also a useful concept called method chaining in Pandas.
In method chaining, we can use more than one method together by connecting them with a dot (.).
The basic structure looks like this:

method1().method2().method3()

Here, the output of the first method becomes the input for the next method.
For example:

marks = pd.Series([85, 65, 95, 72, 90])

marks.sort_values().head(3)

output:

1    65

3    72

0    85

dtype: int64

Here, sort_values() first sorts the marks in ascending order:
65, 72, 85, 90, 95
Then, head(3) returns the first 3 values:
65, 72, 85
This allows us to perform multiple operations in a single line.

• Converting DataFrame Column into a Series
  A DataFrame contains multiple columns, and each individual column of a DataFrame is a Series.
  So, if we want to extract a particular column from a DataFrame as a Series, we can simply select that column.
  For example:

series_marks = df["Marks"]

Here, the Marks column is extracted from the DataFrame and stored in series_marks as a Series.
For example:

series_marks = df["Marks"]

type(series_marks)

output:

pandas.core.series.Series

So, selecting a single DataFrame column using df["Marks"] returns a Series.

• Important Notes

  • Note 1: count() vs size: count() does not count missing values.
    For example:

• marks = pd.Series([80, 90, None, 70])

• marks.size

output:

4

But count() returns only the number of non-null values:

marks.count()

output:

3

So remember:
size → total number of items, including missing values
count() → number of non-null items

• Note 2: isna() and isnull(): isna() and isnull() work in the same way.
  Similarly:
  notna() and notnull() work in the same way.

• Note 3: unique() vs nunique(): unique() returns the actual unique values, while nunique() returns the number of unique values.
  For example:

marks = pd.Series([85, 72, 90, 65, 72])

marks.unique()

output:

[85 72 90 65]

So, unique() returns the unique values.

marks.nunique()

output:

4

So, nunique() returns the number of unique values.

• Note 4: head() vs tail(): head() works from the beginning of the Series, while tail() works from the end of the Series.
  So:
  head() → first few values
  tail() → last few values

• Note 5: sort_values() vs sort_index(): sort_values() sorts the Series according to its values, while sort_index() sorts it according to its index.

• Pandas Series Methods — Short Hand

  • Inspecting: head(), tail(), info(), describe(), unique(), nunique(), value_counts(), memory_usage()

  • Mathematical & Statistical: sum(), mean(), median(), mode(), min(), max(), count(), std(), var(), quantile(), cumsum(), cumprod(), cummax(), cummin()

  • Finding & Checking: isin(), between(), duplicated(), isna(), isnull(), notna(), notnull()

  • Missing Data: dropna(), fillna(), interpolate()

  • Sorting: sort_values(), sort_index(), nlargest(), nsmallest()

  • Transformation: apply(), map(), replace(), astype(), round(), clip()

  • Index & Position: reset_index(), set_axis(), reindex(), take()

  • String Data: str.lower(), str.upper(), str.title(), str.strip(), str.replace(), str.contains(), str.startswith(), str.endswith(), str.len(), str.split()

  • Combining & Comparing: combine(), combine_first(), compare(), equals()

  • Conversion & Export: to_list(), to_numpy(), to_dict(), to_frame(), to_string(), to_csv(), to_json()