how to use pandas filter with IQR

Question

Is there a built-in way to do filtering on a column by IQR(i.e. values between Q1-1.5IQR and Q3+1.5IQR)? also, any other possible generalized filtering in pandas suggested will be appreciated.

Answer 1

As far as I know, the most compact notation seems to be brought by the query method.

# Some test data
np.random.seed(33454)
df = (
    # A standard distribution
    pd.DataFrame({'nb': np.random.randint(0, 100, 20)})
        # Adding some outliers
        .append(pd.DataFrame({'nb': np.random.randint(100, 200, 2)}))
        # Reseting the index
        .reset_index(drop=True)
    )

# Computing IQR
Q1 = df['nb'].quantile(0.25)
Q3 = df['nb'].quantile(0.75)
IQR = Q3 - Q1

# Filtering Values between Q1-1.5IQR and Q3+1.5IQR
filtered = df.query('(@Q1 - 1.5 * @IQR) <= nb <= (@Q3 + 1.5 * @IQR)')

Then we can plot the result to check the difference. We observe that the outlier in the left boxplot (the cross at 183) does not appear anymore in the filtered series.

# Ploting the result to check the difference
df.join(filtered, rsuffix='_filtered').boxplot()

Since this answer I've written a post on this topic were you may find more information.

Answer 2

Another approach using Series.between():

iqr = df['col'][df['col'].between(df['col'].quantile(.25), df['col'].quantile(.75), inclusive=True)]

Drawn out:

# Select the first quantile
q1 = df['col'].quantile(.25)

# Select the third quantile
q3 = df['col'].quantile(.75)

# Create a mask inbeetween q1 & q3
mask = df['col'].between(q1, q3, inclusive=True)

# Filtering the initial dataframe with a mask
iqr = df.loc[mask, 'col']
        

Answer 3

This will give you the subset of df which lies in the IQR of column column:

def subset_by_iqr(df, column, whisker_width=1.5):
    """Remove outliers from a dataframe by column, including optional 
       whiskers, removing rows for which the column value are 
       less than Q1-1.5IQR or greater than Q3+1.5IQR.
    Args:
        df (`:obj:pd.DataFrame`): A pandas dataframe to subset
        column (str): Name of the column to calculate the subset from.
        whisker_width (float): Optional, loosen the IQR filter by a
                               factor of `whisker_width` * IQR.
    Returns:
        (`:obj:pd.DataFrame`): Filtered dataframe
    """
    # Calculate Q1, Q2 and IQR
    q1 = df[column].quantile(0.25)                 
    q3 = df[column].quantile(0.75)
    iqr = q3 - q1
    # Apply filter with respect to IQR, including optional whiskers
    filter = (df[column] >= q1 - whisker_width*iqr) & (df[column] <= q3 + whisker_width*iqr)
    return df.loc[filter]                                                     

# Example for whiskers = 1.5, as requested by the OP
df_filtered = subset_by_iqr(df, 'column_name', whisker_width=1.5)

Answer 4

Find the 1st and 3rd quartile using df.quantile and then use a mask on the dataframe. In case you want to remove them, use no_outliers and invert the condition in the mask to get outliers.

Q1 = df.col.quantile(0.25)
Q3 = df.col.quantile(0.75)
IQR = Q3 - Q1
no_outliers = df.col[(Q1 - 1.5*IQR < df.BMI) &  (df.BMI < Q3 + 1.5*IQR)]
outliers = df.col[(Q1 - 1.5*IQR >= df.BMI) |  (df.BMI >= Q3 + 1.5*IQR)]

Answer 5

Another approach uses Series.clip:

q = s.quantile([.25, .75])
s = s[~s.clip(*q).isin(q)]

here are details:

s = pd.Series(np.randon.randn(100))
q = s.quantile([.25, .75])  # calculate lower and upper bounds
s = s.clip(*q)  # assigns values outside boundary to boundary values
s = s[~s.isin(q)]  # take only observations within bounds

Using it to filter a whole dataframe df is straightforward:

def iqr(df, colname, bounds = [.25, .75]):
    s = df[colname]
    q = s.quantile(bounds)
    return df[~s.clip(*q).isin(q)]

Note: the method excludes the boundaries themselves.

Answer 6

You can try using the below code, also, by calculating IQR. Based on the IQR, lower and upper bound, it will replace the value of outliers presented in each column. this code will go through each columns in data-frame and work one by one by filtering the outliers alone, instead of going through all the values in rows for finding outliers.

Function:

    def mod_outlier(df):
        df1 = df.copy()
        df = df._get_numeric_data()


        q1 = df.quantile(0.25)
        q3 = df.quantile(0.75)

        iqr = q3 - q1

        lower_bound = q1 -(1.5 * iqr) 
        upper_bound = q3 +(1.5 * iqr)


        for col in col_vals:
            for i in range(0,len(df[col])):
                if df[col][i] < lower_bound[col]:            
                    df[col][i] = lower_bound[col]

                if df[col][i] > upper_bound[col]:            
                    df[col][i] = upper_bound[col]    


        for col in col_vals:
            df1[col] = df[col]

        return(df1)

Function call:

df = mod_outlier(df)