Alternative to np.where when requiring numerous changes to a column in a DataFrame

Asked Jul 07 '21 at 19:22

Active Jul 07 '21 at 19:22

Viewed 35 times

In Python, if I have a data frame in which I want to alter the contents in a particular column from one specific value to another, I so far use the following method using numpy:

df["Country"] = np.where(df["Country"] == "UK", "United Kingdom", df["Country"])
df["Country"] = np.where(df["Country"] == "USA", "United States", df["Country"])

I have about 100 of these types of changes, which looks ungainly in code. Can an example be provided in how to alter the above code in a way that would look more elegant when having numerous extra changes?

asked Jul 07 '21 at 19:22

user1309044

You could replace with a `dict` -> `df['Country'] = df['Country'].replace({'UK': 'United Kingdom', 'USA': 'United States'})` – Henry Ecker Jul 07 '21 at 19:30
See [Remap values in pandas column with a dict](https://stackoverflow.com/q/20250771/15497888) – Henry Ecker Jul 07 '21 at 19:30
how about a merge / join to another dataframe that soft encodes your mappings? – Rob Raymond Jul 07 '21 at 20:09

Alternative to np.where when requiring numerous changes to a column in a DataFrame

0 Answers0