Scatter plot of categorical data from multiple data frame columns

Question

I have a DataFrame df like this:

   name         Int1        Int2         Int3
0     foob -2.534519e-05 -1.156744e-04 -1.410195e-04
1     arz  2.907239e-04  3.502863e-04  6.410102e-04
2     foo2 -2.140769e-04  4.214626e-04  2.073857e-04
3     bar  3.366116e-03 -6.125303e-04  2.753586e-03
4     rnd -5.014413e-05 -6.740579e-06 -5.688471e-05
5     baz  3.334906e-04 -7.846232e-05  2.550283e-04
6     rnd2 -6.111069e-04  2.194443e-03  1.583336e-03
7     tet  3.184057e-04  2.208398e-04  5.392455e-04

The df should be plotted as a scatter plot depicting three data points (Int1, Int2, Int3) per name.

Currently, I am using seaborn's stripplot function, which works fine if I assign a plot of each single column (e.g. x=name, y=Int1) to the same axis of a figure:

fig, ax = plt.subplots()
seaborn.stripplot(df.name, df.Int1, ax=ax, c='red')
seaborn.stripplot(df.name, df.Int2, ax=ax, c='blue')

However, I want to plot this in a better way for the main reason of having a proper legend and better customization. The solution can also be pandas based.

I take it from your example that you want the name to appear on the horizontal axis, and above each one should be three points representing the three values from the row with that name? Do you want the points in each column to be the same color? If not, do you want a consistent color scheme that uses a single color for Int1, a different color for Int2, and another color for Int3? — David Z
– David Z, Commented Jan 2, 2017 at 10:14
@DavidZ Fully correct. The color scheme should be consistent among a single column (Int1) and different between each column (Int1 vs Int2 vs Int3). E.g. red, green, and blue for Int1, Int2, and Int3, respectively. Thanks. — Fourier
– Fourier, Commented Jan 2, 2017 at 10:17

Fourier · Accepted Answer · 2017-01-02 12:00:10Z

2

Here is my solution. Actually quite simple:

df_melt=pd.melt(df,id_vars=['name'], var_name='intensities', value_name='values')
sns.stripplot(x="name", y="values", data=df_melt, hue='intensities')

This takes the original df and with the help of pandas melt function, produces a new df containing a single intensities column with one row for each of the Int1, Int2, and Int3 values per name. The second line uses seaborn's stripplot to plot df_melt colored by the respective column Int1, Int2, and Int3

answered Jan 2, 2017 at 12:00

Fourier

3,0033 gold badges29 silver badges40 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Scatter plot of categorical data from multiple data frame columns

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related