Hence, having all variables on the same scale will facilitate easy comparison of the “importance” of each variable, as now all variables are on the same scale. The most common way to standardize the variable X X is to use the z z transformation: zi = xi −μ sdX z i = x i − μ s d X. Yes, I realize there is a multicollinearity problem. Let's ignore that for the moment. I can run a regression on this data (using R) using a standard lm() command. If you know the min and max before scaling, then yes. If you don't, then no. It's unclear what you're looking to do. You could reindex the initial frame based on the remaining values in the cleansed frame. Sample data, source code, and expected output would be helpful to give a more detailed answer. I guess you means the inverse_transform What I've done first, is rescaled the data using min-max normalization: # Normalize data between 0 and 1 from sklearn.preprocessing import MinMaxScaler min_max = MinMaxScaler() dataframe2 = pd.DataFrame(min_max.fit_transform(dataframe), columns = dataframe.columns) {"payload":{"allShortcutsEnabled":false,"fileTree":{"":{"items":[{"name":".RData","path":".RData","contentType":"file"},{"name":".Rhistory","path":".Rhistory Scaled data is only for the machine learning methods that need well-conditioned data for processing. Once the training or prediction is completed, the data needs to be returned to the unscaled form for visualization or interpretation. The inverse_transform function is used to unscale the data. x= y a +b x = y a + b. R unscale and back transform plot axis or use axis from original data column. I am plotting a variable's effect on a modeled fit. The variable was sqrt transformed and then scaled. I can plot the original values of 'weight' against the modeled fit but the resulting geom_line is very different and the range on the x-axis where the large increase Philipp Schmid. Earlier this year, Google introduced and open sourced FLAN-T5, a better T5 model in any aspect. FLAN-T5 outperforms T5 by double-digit improvements for the same number of yp0bx.