You shouldn't be asking this type of question on Stackoverflow. Instead, please post what you've done/tried and what error has surfaced while doing so. This way you would learn about the problem much faster instead of just asking for pure answers. Please refrain from repeating this mistake in the future. My take on this problem, however, would be something like this:
cols = df.columns
most_freq = []
for col in cols:
word = df[col].value_counts().idxmax()
most_freq.append(word)
word_series = pd.Series(most_freqm index=cols)
df = df.append(word_series, ignore_index=True)
Explanation:
I would use value_counts().idxmax()
to find the most frequent word in each column and loop through all the columns in the dataframe. I would then put those most-frequent-words into a list which I will convert into a series. This series would then be appended into the original dataframe.
Cheers! Happy coding!