CODEWITHSUNDEEP
pythonpandas
Chad Kennedy
Chad Kennedy

Reputation: 1736

Add fractional component in Pandas

I have a dataframe with a column with values like the following:

1 1 1 1 2 2 2 2 3 3 3 3 etc

I would like to change the values to

1.0 1.25 1.5 1.75 2.0 2.25 2.5 2.75 3.0 3.25 3.5 3.75 etc

The initial integer values are always monotonically increasing, but may have gaps. They will always be repeated exactly 4 times.

I have implemented this via a for loop, but this takes a long time to run on a large data set. I'm looking for a more efficient way.

for i in range(len(df) // 4):
    for j in range(4):
        df.timestamp.iloc[i * 4 + j] += j / samples_per_sec

Upvotes: 2

Views: 63

Answers (2)

EdChum
EdChum

Reputation: 394179

You could do it this way:

In [47]:
l=[1, 1, 1, 1, 2, 2, 2, 2 ,3 ,3 ,3, 3]
df = pd.DataFrame({'values':l})
df['values'] = df['values'] + (0.25 * (df.index.values % 4 ))
df

Out[47]:
    values
0     1.00
1     1.25
2     1.50
3     1.75
4     2.00
5     2.25
6     2.50
7     2.75
8     3.00
9     3.25
10    3.50
11    3.75

So assuming that any values that are present are always repeated 4 times, as you've stated then the above should work.

using another dataset with gaps:

In [48]:
l=[1, 1, 1, 1, 2, 2, 2, 2 ,4 ,4 ,4, 4,7,7,7,7]
df = pd.DataFrame({'values':l})
df['values'] = df['values'] + (0.25 * (df.index.values % 4 ))
df

Out[48]:
    values
0     1.00
1     1.25
2     1.50
3     1.75
4     2.00
5     2.25
6     2.50
7     2.75
8     4.00
9     4.25
10    4.50
11    4.75
12    7.00
13    7.25
14    7.50
15    7.75

Upvotes: 2

Sebastien
Sebastien

Reputation: 153

You could do something like this

df.timestamp += [(j % samples_per_sec)*1. / samples_per_sec for j in range(len(df))]

Note: I'm assuming that samples_per_sec = 4.

Upvotes: 1

Related Questions