This method is used for label-based indexing. For example, 5 of 'a'. Notice that in this case 5 is a label of and index not a location. 
An array or list of labels can be passed to retrieve multiple rows, ['a', 'b', 'c']


Pandas DataFrame Indexing .loc

.iloc is integer position based, so you must use the row or column integer position to access the data.


Given the pictured dataframe, 


If you want to access all of the “viper” values you would run... 
```python
df.iloc[1]
```
This would return a Pandas Series of the “viper” row.


If you want to extract multiple rows, you can include the integer values in a list like... 
```python
df.iloc[[0,1]]
```
This returns a Pandas Dataframe of all of the values from both the “cobra” and “viper” rows.


To get a single value such as the max_speed of a sidewinder, you can use...
```python
df.iloc[2,0]
```
This would return a single integer value of 7. 


The syntax for the above example is... 
```python
df.iloc[row_number, column_number]
``` 

Remember that Python is a zero-based indexing language so a row or column in the first position has a row or column number of 0 and not 1.


Pandas DataFrame .iloc Slicing

Pandas DataFrames support label-based indexing with `.loc` and integer-position-based indexing with `.iloc`. For example, `df.iloc[3]` selects the row at integer position 3, while `df.iloc[3:5, 0:2]` selects rows at positions 3 and 4 and the first two columns. `df.loc[dates[0]]` selects a labeled row, and `df.loc[:, ["A", "B"]]` selects the labeled columns A and B. Boolean conditions can filter data: `df[df["A"] > 0]` returns rows whose value in column A is positive. The related `.at` and `.iat` accessors also provide direct data access. Assignment, such as `df["A"] = 0`, modifies column A in the DataFrame; this should not be described as meaning that every data access is pass-by-reference.

Google

University of Michigan - Ann Arbor

A Pandas DataFrame can be subset by column names, row labels, integer positions, or Boolean conditions. Use `df["column_name"]` to select one column. For explicit two-dimensional selection, `.loc` uses labels and `.iloc` uses integer positions; for example, `df.iloc[3:5, 0:2]` selects rows at positions 3 and 4 and the first two columns. A label slice such as `df["20130102":"20130104"]` selects rows by index labels, while a condition such as `df[df["A"] > 0]` filters rows according to column values.

Learn Before

Related

Learn After