大熊猫自动增量指数

Question

How can I auto increment indexes in pandas and even if I delete the existing record then while appending new record the indexes should not get reallocated or the index once alloted to a record should not change? I have tried append with both of its options and it changes the index. Feature somewhat like the MySQL AUTO_INCREMENT will be of great help.

Answer 1

Pandas is not a database. Deleting and inserting single rows has disastrous performance. Instead, remove rows by masking them out:

maxidx = df.index[-1]
mask = [True, False, True] # etc
df = df[mask]

The retained rows will keep their original index values.

Then to add multiple rows at once:

to_add = # some dataframe
to_add.index = pd.RangeIndex(maxidx + 1, maxidx + 1 + len(to_add))
df = pd.concat([df, to_add])