Assignment with pandas iloc where RHS length is shorter than LHS-CodePudding

I am trying to perform np.where function on a dataframe starting from row 20 onward. The code that I entered as follow:

df['buy'] = np.where((df.iloc[20:,]['signal']==1), 'buy','no buy')

It showed the error below: ValueError: Length of values (226) does not match length of index (246)

Anyone know how to fix it please?

I want pandas to return value from row 20 onward.

CodePudding user response：

This is an assignment error, df has 246 entries, but you are trying to assign a numpy vector with 226 values to it.

What do you intend the result should be?

you could subset the dataframe?

df_sub = df.iloc[20:].copy()
df_sub['buy']=np.where((df_sub.iloc['signal']==1), 'buy','no buy')

CodePudding user response：

IIUC, you could use "mixed" indexing for the assignment:

n = df.columns.get_loc('buy')
df.iloc[20:, n] = np.where((df.iloc[20:,]['signal']==1), 'buy','no buy')