What you should see here is that there appears to be some general positive relationship but note that there are outliers in circulation, where circulation is much higher than number of Pulitzers won as well as outliers in Pulitzers won. Overall, most of the data are clustered around zero Pulitzers. This is because the distributions of both of our variables are skewed right. This suggests that we may need to transform these data to use linear regression to ensure that the large outlier values are not driving our relationship. So, as discussed in class: Apply a log10-transformation the Pulitzer count data, with an offset of 0.1 This will shift the values away from being centered near zero when put on the lo scale. Store this in a new column pulitzer_log10. [40]: pulitzer ['pulitzer_log10'] = np. log10(pulitzer['Pulitzer Prize Winners and_ Finalists, 1990-2014'] + 0.1) [43] : sns . histplot (pulitzer ['pulitzer_log10' ], kde=True, bins=30) ax1 . set_xlabel (' 1990-2003' ) axl . set_ylabel ('Number of Pulitzers') NameError Traceback (most recent cally last) in 2 kde=True, bins=30) 3 ax1 . set_xlabel (' 1990-2003' ) 4 axl. set_ylabel ('Number of Pulitzers' ) NameError: name 'axl' is not defined un Count N 0 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 pulitzer_log10