The Basics of Timing: Selecting Columns and Rows in Pandas DataFrames
In our previous lesson, we explored the basics of timing how to time a chunk of code and why speed efficiency matters. In this lesson, we will dive deeper into the pandas library and focus on selecting columns and rows in data frames. Specifically, we will be using the `iloc` function and compare its performance with the `loc` function.
One of the most useful features of the pandas library is the ease of convenience of selecting specific rows of a panda's data frame. We can use the index number locator or the index name locator to target rows in our data frame. In this example, we want to select the first 500 rows of the poker data set. We will first use the `iloc` function and then use the `loc` function to achieve the same result.
The syntax for using `iloc` is straightforward: we simply specify the row index that we want to start from and end at. In this case, we want to select the first 500 rows, so we would use `dataframe.iloc[0:500]`. However, it's worth noting that `iloc` takes advantage of the order of the indices which are already sorted, making it faster than using `loc`.
On the other hand, the syntax for using `loc` is also simple: we specify a label selector and a boolean mask. In this case, we want to select the first 500 rows, so we would use `dataframe.loc[0:500]`. While both methods have the same syntax, `iloc` performs almost 200% faster than `loc`.
Another important feature of the pandas library is its ability to select specific columns in a data frame. We can do this by using the `loc` function or by specifying column names directly. In this example, we want to select the first three columns of the poker data set: the symbol, the rank of the cards that came first in each hand, and the rank of the cards that came second in each hand.
To achieve this, we can use the `loc` function with an integer position or a label selector. For example, we could use `dataframe.loc[:, ['symbol', 'rank_first', 'rank_second']]`. Alternatively, we can simply specify the column names by including them in double square brackets: `dataframe[["symbol", "rank_first", "rank_second"]]`.
In terms of speed, it's worth noting that `loc` performs 30% faster than `iloc` when it comes to selecting columns. This is because `loc` can take advantage of the column data type being sorted alphabetically, making it more efficient.
In summary, both `iloc` and `loc` are useful functions for selecting rows and columns in pandas data frames, but they have different strengths and weaknesses depending on the specific task at hand. By understanding how to use these functions effectively, we can improve the performance of our code and make it more efficient.
"WEBVTTKind: captionsLanguage: enwelcome back in the previous lesson we studied the basics of timing how to time a chunk of code and why speed efficiency matters in general in this lesson we will look at the log and I love pandas function and find out which one is the most efficient to select columns and rows in the pandas dataframe let's look at the main data set we will use in this lesson which derived from the famous poker card game in each round its player has 5 cards in hand each one characterized by its symbol which can be either hard diamonds clubs or spades Eddie trunk which ranges from 1 to 13 the data set consists of every possible combination of five cards one person can possess let's take for example the first combination which correspond to the first row we have a ten of diamonds a jack and the king of clubs and four spades and an Ace of Hearts if you're still not completely sure about the data set please pause the video and look the bottom part of this slide carefully one of the most useful features of the palace library is the ease of convenience of selecting specific rows of a panda's data frame we're going to use a lock the index number locator and lock the index name locator in this example we want to select the first 500 rows of the poker data set firstly by using the lock function and then by using the eye log function while these two methods have the same syntax ILOG performs almost 200% faster than lock a lock takes advantage of the order of the indices which are already sorted and is therefore faster we use a lock and lock to target rows but we can also use them to locate different features in the pandas dataframe in this example we want to select the first three columns of the poker data set the symbol and the rank of the cards that came first in each hand and the rank of the cards that came second in each hand we can use the iya log function to locate a feature by index alternatively we can simply select one or several columns by name the syntax of a log for that purpose is simple we denote with a column that we want all the rows of the data frame and then after the comma we use a colon followed by a 3 to denote that we want all the columns until the third one to select columns by name we simply include the name of the columns we want in double square brackets in terms of speed for the task of locating features lock performs 30% faster in general the lock function works better for selecting columns while dialogue is faster for selecting specific rows now that we explore the differences between a log and log it's your turn to target rows and columns and evaluate thewelcome back in the previous lesson we studied the basics of timing how to time a chunk of code and why speed efficiency matters in general in this lesson we will look at the log and I love pandas function and find out which one is the most efficient to select columns and rows in the pandas dataframe let's look at the main data set we will use in this lesson which derived from the famous poker card game in each round its player has 5 cards in hand each one characterized by its symbol which can be either hard diamonds clubs or spades Eddie trunk which ranges from 1 to 13 the data set consists of every possible combination of five cards one person can possess let's take for example the first combination which correspond to the first row we have a ten of diamonds a jack and the king of clubs and four spades and an Ace of Hearts if you're still not completely sure about the data set please pause the video and look the bottom part of this slide carefully one of the most useful features of the palace library is the ease of convenience of selecting specific rows of a panda's data frame we're going to use a lock the index number locator and lock the index name locator in this example we want to select the first 500 rows of the poker data set firstly by using the lock function and then by using the eye log function while these two methods have the same syntax ILOG performs almost 200% faster than lock a lock takes advantage of the order of the indices which are already sorted and is therefore faster we use a lock and lock to target rows but we can also use them to locate different features in the pandas dataframe in this example we want to select the first three columns of the poker data set the symbol and the rank of the cards that came first in each hand and the rank of the cards that came second in each hand we can use the iya log function to locate a feature by index alternatively we can simply select one or several columns by name the syntax of a log for that purpose is simple we denote with a column that we want all the rows of the data frame and then after the comma we use a colon followed by a 3 to denote that we want all the columns until the third one to select columns by name we simply include the name of the columns we want in double square brackets in terms of speed for the task of locating features lock performs 30% faster in general the lock function works better for selecting columns while dialogue is faster for selecting specific rows now that we explore the differences between a log and log it's your turn to target rows and columns and evaluate the\n"
