This expression
df["column"].diff() != 0
Gives me a pandas series of booleans. I now would like a column df["result"] where there's the value 100 for every True value in df["column"] and 0 for every False.
I don't understand why this doesn't work:
df["result"] = 100 if df["column"].diff() != 0 else 0
I understand I'd have to use loc, but from this:
df.loc[df["column"].diff() != 0]
How do I then set the result column?
Here is best use numpy.where for set 2 values by condition, solution is vectorized:
df["result"] = np.where(df["column"].diff() != 0, 100, 0)
Your code:
df["result"] = 100 if df["column"].diff() != 0 else 0
not working, because here are used 1d arrays, (Series) so cannot be used scalar solution.
You also get:
ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().
And it means there is no clear, what is scalar output from array of boolean from df["column"].diff() != 0.
More information is in Using if/truth statements with pandas
May be you are looking for something like this:
df.loc[df['column'].diff()!=0, 'Result'] = 100
Edit after jezrael's comment:
df['diff'] = df['column'].diff().fillna(0)
df.loc[df['diff'] != 0, 'Result'] = 100
