I am kind of getting stuck on extracting value of one variable conditioning on another variable. For example, the following dataframe:
A B
p1 1
p1 2
p3 3
p2 4
How can I get the value of A when B=3? Every time when I extracted the value of A, I got an object, not a string.
You could use loc to get series which satisfying your condition and then iloc to get first element:
In [2]: df
Out[2]:
A B
0 p1 1
1 p1 2
2 p3 3
3 p2 4
In [3]: df.loc[df['B'] == 3, 'A']
Out[3]:
2 p3
Name: A, dtype: object
In [4]: df.loc[df['B'] == 3, 'A'].iloc[0]
Out[4]: 'p3'
df.loc[df['B'] == 3, 'A'].item() works for me too.
You can try query, which is less typing:
df.query('B==3')['A']
query creates a new DataFrame, and then the slice ['A'] creates yet another DataFrame. For large cases one would like to avoid the unnecessary intermediate object.Try:
df[df['B']==3]['A'].item()
assuming df is your pandas.DataFrame.
][ part.
Use df[df['B']==3]['A'].values[0] if you just want item itself without the brackets
Edited: What I described below under Previous is chained indexing and may not work in some situations. The best practice is to use loc, but the concept is the same:
df.loc[row, col]
row and col can be specified directly (e.g., 'A' or ['A', 'B']) or with a mask (e.g. df['B'] == 3). Using the example below:
df.loc[df['B'] == 3, 'A']
Previous: It's easier for me to think in these terms, but borrowing from other answers. The value you want is located in a dataframe:
df[*column*][*row*]
where column and row point to the values you want returned. For your example, column is 'A' and for row you use a mask:
df['B'] == 3
To get the first matched value from the series there are several options:
df['A'][df['B'] == 3].values[0]
df['A'][df['B'] == 3].iloc[0]
df['A'][df['B'] == 3].to_numpy()[0]
You can use squeeze instead of iloc[0]. It looks clearer if you have only one value:
df.loc[df['B'] == 3, 'A'].squeeze()
Output:
'p3'
df.loc[df['B']=='give-specific-value', 'A']
I have also worked on this clausing and extraction operations for my assignment.
item()at the end.df.queryandpd.evalseem like good fits for this use case. For information on thepd.eval()family of functions, their features and use cases, please visit Dynamic Expression Evaluation in pandas using pd.eval().