This should work, can you show what df_cmp looks like? and what sort of object it is using 

@emma_s , here is what you asked:

type(df_cmp) :

pyspark.sql.classic.dataframe.DataFrame

df_cmp.describe():

DataFrame[
run_summary: string,
entity_id: string,
location_id: string,
category_id: string,
elasticity_value: string,
demand_forecast: string,
execution_flag: string,
execution_status: string,
elasticity_diff: string,
elasticity_rel_diff: string,
elasticity_diff_signed: string,
forecast_diff: string,
forecast_rel_diff: string
]

 Image of some values:

Why the "dt" value is not shown when using the describe method?

The describe looks like a completely different table to the image of values, they have different field names and different types. At what point do you get the error? Does it let you get the visualisation pop up? It should work fine. Is it when you try to put the histogram on?