LM - Linear Model plot

seaborn.lmplot() plots data points (scatter) along with a linear regression model fit, just like sns.regplot().
However, unlike regplot(), lmplot() is a figure-level function, meaning it can create:

• Multiple subplots (facets) based on categorical variables
• Grouped regressions using color (hue)

It’s built on top of sns.regplot() and adds faceting (via FacetGrid) for multi-group or multi-category analysis.

Syntax:

Parameters:

• data = DataFrame containing the data
• x, y = Numeric variables for regression
• hue = Adds subgroups (different colours and regression lines)
• col, row = Facet the data into multiple subplots
• palette = Colour palette for hue groups
• col_wrap = Wrap facet columns into multiple rows
• height = Height (in inches) of each facet
• aspect = Aspect ratio (width/height) of each facet
• order = Degree of polynomial regression (1 = linear)
• robust = Use robust regression to reduce outlier influence
• ci = Confidence interval width (in %)
• line_kws = Keyword args for line (colour, style, width, etc.)
• scatter_kws = Keyword args for scatter points
• legend = Show/hide legend
• legend_out = Place legend outside the plot grid
• facet_kws = Pass additional parameters to the FacetGrid

Basic example

Plots a scatterplot with a regression line and 95% confidence interval. Figure-level → creates its own figure and axes.

Facet into multiple columns

Creates one subplot per day. Each shows its own scatter + regression line.

Facet by rows and columns

Creates a grid of subplots, split by both day (columns) and sex (rows).

Wrap columns across rows

Creates 2 columns per row (better layout for many facets).

Compare groups side-by-side

Multiple regression lines across columns for day, coloured by sex.