Subplots using differnet columns?

[ziliangok]

For example:
plot1: px.box(df, y='feature1')
plot2: px.box(df, y='feature2')
The facet_row and facet_col can make subplots on one columns, but I want to make subplots on differents columns.

[nicolaskruchten]

Plotly Express operates on "tidy data" only, so you will need to reshape your data frame before passing it into px.box. This is pretty easy with the melt() function in Pandas:

import pandas as pd
df = pd.read_csv(pd.compat.StringIO("""
PlantID,A,B,C,D
1,0,1,2,4
1,3,0,2,0
3,0,0,0,1
4,0,1,1,5
"""))
print(df.head())

   PlantID  A  B  C  D
0        1  0  1  2  4
1        1  3  0  2  0
2        3  0  0  0  1
3        4  0  1  1  5

print(df.melt(id_vars=["PlantID"]).head())

   PlantID variable  value
0        1        A      0
1        1        A      3
2        3        A      0
3        4        A      0
4        1        B      1

Meaning that you can do:

import plotly.express as px
px.box(df.melt(id_vars=["PlantID"]), x="variable", y="value")

[nicolaskruchten]

Ah sorry, if you want facets per original column it's just:

import plotly.express as px
px.box(df.melt(id_vars=["PlantID"]), y="value", facet_col="variable", boxmode="overlay")

[ziliangok]

Thank you! Although this is not what I want, it solve my problems. (●＾o＾●)

[nicolaskruchten]

We've actually implemented support for subplots per columns! https://medium.com/plotly/beyond-tidy-plotly-express-now-accepts-wide-form-and-mixed-form-data-bdc3e054f891