I have dataframe like this
df = pd.DataFrame({'_3321131460': ['col1', 'col1', 'col2'], '_3952604542': ['col1', 'col2', 'col2']})
df
So want to create graph like this
one graph but x are two columns and legent have two columns data
And y is count But know how to do that
So one columns histogram, another columns histogram in one graph
But have same legend
You can transform your df with melt in order to have the column names in their own column:
import pandas as pd
import plotly.express as px
df = pd.DataFrame({'_3321131460': ['col1', 'col1', 'col2'], '_3952604542': ['col1', 'col2', 'col2']})
df_t = df.melt()
This will have the following form (feel free to rename the columns to get fitting labels on your final plot):
variable value
0 _3321131460 col1
1 _3321131460 col1
2 _3321131460 col2
3 _3952604542 col1
4 _3952604542 col2
5 _3952604542 col2
Then use px.bar:
fig = px.bar(df_t, x='variable', color='value', barmode="group")
fig.show()
Output:
