CODEWITHSUNDEEP
pythonpandascsv
pablitoo
pablitoo

Reputation: 25

Delete last rows using a condition

I would like to delete the last rows based on a condition. For example I have the following columns :

  voltage | Current
0      10 | 0.8
1      12 | 0.7
3      14 | 0.6
4       0 | -0.0001
5      10 | 0.8
6      12 | 0.7
7      14 | 0.6
8       0 | -0.0001
9       0 | -0.0001

In this case, I want to remove the 2 last rows (when voltage = 0) without removing the 4th row.

I was thinking about a while loop that starts from the end of the dataframe and delete all rows with voltage = 0. It will stop when voltage is different from 0.

Any idea ?

Upvotes: 1

Views: 518

Answers (4)

Dodd-learning
Dodd-learning

Reputation: 118

There are some good answers already but I saw all have some potential faults (at the time of posting). I have provided a solution that is a simple and fast filter.

Here we are saying:

Take the first index_filter number of rows regardless of value to capture any voltages you want to keep 0.

Then, we filter out anything where voltage == 0 (excluding the index_filter rows).

import pandas as pd
data = [[10, 0.8], [0, -0.0001], [12, 0.7], [0, -0.0001], [0, -0.0001]]
df = pd.DataFrame(columns=['voltage', 'Current'], data=data)
INDEX_FILTER = 1  # remember, index starts at 0 and we include this in our filter
df = df.loc[((df.index <= INDEX_FILTER) | (df['voltage'] != 0))]

I have used something similar in my electrical engineering labs to filter out many op amp circuits :).

Thanks

Upvotes: 2

Ovski
Ovski

Reputation: 662

Your idea of a reverse loop works just fine.

values = [
    (10, 0.8),
    (12, 0.7),
    (14, 0.6),
    (0, -0.0001),
    (10, 0.8),
    (12, 0.7),
    (14, 0.6),
    (0, -0.0001),
    (0, -0.0001),
]

print(values)
for i in range(len(values) - 1, -1, -1):
    if values[i][0] == 0:
        del(values[i])
    else:
        break
print(values)

Output:

[(10, 0.8), (12, 0.7), (14, 0.6), (0, -0.0001), (10, 0.8), (12, 0.7), (14, 0.6), (0, -0.0001), (0, -0.0001)]

[(10, 0.8), (12, 0.7), (14, 0.6), (0, -0.0001), (10, 0.8), (12, 0.7), (14, 0.6)]

Upvotes: 2

Andrej Kesely
Andrej Kesely

Reputation: 195438

Try:

df = df[df["voltage"].replace(0, np.nan).bfill().notna()]
print(df)

Prints:

   voltage  Current
0       10   0.8000
1       12   0.7000
2       14   0.6000
3        0  -0.0001
4       10   0.8000
5       12   0.7000
6       14   0.6000

Upvotes: 2

JNevill
JNevill

Reputation: 50034

You can use .loc to reverse the dataframe and then grab the .idxmax() where voltage > 0. Then use loc once again to keep everything up to that idxmax()

df = pd.DataFrame({"voltage":[10,12,14,0,10,12,14,0,0], "Current":[1,2,3,4,5,6,7,8,9]})
 
df.loc[:((df.loc[::-1]['voltage'] > 0) ).idxmax()]

+-------+---------+---------+
| index | voltage | Current |
+-------+---------+---------+
|     0 |      10 |       1 |
|     1 |      12 |       2 |
|     2 |      14 |       3 |
|     3 |       0 |       4 |
|     4 |      10 |       5 |
|     5 |      12 |       6 |
|     6 |      14 |       7 |
+-------+---------+---------+

Upvotes: 2

Related Questions