stackoom
  简体   繁体   中英

How to replace certain values in a column based on a certain condition?

 原文  2019-05-02 12:56:31  3   5    python/ python-3.x/ pandas

Question

I have the following df

col_1   col_2
1   1
1   2
1   3
1   6
1   8
1   11
1   12
1   19
1   24
1   1
1   1
1   2
1   2
1   3 
1   3
2   1
2   2
2   4
2   6
2   7
2   11
2   13
2   16
2   19
2   1
2   2
2   3

I would like to do kind of groupby on col_1 and replace the values 1, 2, 3 that occur after 19 in the col_2 and change them with 25, 26, 27.

Expected Output:

col_1   col_2
1   1
1   2
1   3
1   6
1   8
1   11
1   12
1   19
1   24
1   25
1   25
1   26
1   26
1   27
1   27
2   1
2   2
2   4
2   6
2   7
2   11
2   13
2   16
2   19
2   25
2   26
2   27

I would like to know how can this be done using pandas.

Thanks

Edit 1:

My real df

ContextID   BacksGas_Flow_sccm  StepID
7289973 1.953125    1
7289973 2.05078125  2
7289973 2.05078125  2
7289973 2.05078125  2
7289973 1.953125    2
7289973 1.7578125   2
7289973 1.7578125   2
7289973 1.85546875  2
7289973 1.7578125   2
7289973 9.08203125  5
7289973 46.19140625 5
7289973 46.19140625 5
7289973 46.19140625 5
7289973 46.19140625 5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  5
7289973 46.6796875  7
7289973 46.6796875  7
7289973 46.6796875  7
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  12
7289973 46.6796875  15
7289973 46.6796875  15
7289973 46.6796875  16
7289973 46.6796875  16
7289973 46.6796875  17
7289973 25.09765625 19
7289973 45.99609375 19
7289973 59.08203125 19
7289973 61.81640625 19
7289973 62.59765625 19
7289973 63.671875   19
7289973 65.625  19
7289973 66.69921875 19
7289973 67.3828125  19
7289973 67.3828125  19
7289973 67.67578125 19
7289973 68.26171875 19
7289973 69.04296875 19
7289973 69.82421875 19
7289973 69.82421875 19
7289973 70.8984375  19
7289973 70.8984375  19
7289973 70.8984375  19
7289973 70.8984375  19
7289973 71.6796875  19
7289973 71.6796875  19
7289973 72.55859375 19
7289973 72.55859375 19
7289973 72.55859375 19
7289973 72.55859375 19
7289973 72.55859375 19
7289973 72.55859375 19
7289973 73.33984375 19
7289973 73.33984375 19
7289973 73.33984375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 73.2421875  19
7289973 73.2421875  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.0234375  19
7289973 74.90234375 19
7289973 74.90234375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 75          19
7289973 75          19
7289973 75          19
7289973 74.21875    19
7289973 74.21875    19
7289973 74.21875    19
7289973 75          19
7289973 75          19
7289973 75          19
7289973 75          19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.12109375 19
7289973 74.90234375 19
7289973 6.4453125   24
7289973 3.515625    24
7289973 2.5390625   24
7289973 2.05078125  24
7289973 2.05078125  24
7289973 2.05078125  24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.953125    24
7289973 2.05078125  24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 2.05078125  24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.7578125   24
7289973 1.66015625  24
7289973 1.7578125   24
7289973 1.7578125   24
7289973 1.7578125   24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.953125    24
7289973 1.7578125   24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.7578125   24
7289973 1.85546875  24
7289973 1.85546875  24
7289973 1.7578125   24
7289973 1.7578125   1
7289973 1.85546875  1
7289973 1.85546875  1
7289973 1.85546875  2
7289973 1.7578125   2
7289973 1.953125    2
7289973 1.953125    2
7289973 1.85546875  2
7289973 1.85546875  3
7289973 1.85546875  3
7289973 1.85546875  3
7289973 1.953125    3
7289973 1.85546875  3
7289973 1.953125    3
7289973 1.85546875  3
7289973 1.7578125   3
7289973 1.85546875  3
7289973 1.85546875  3
7289973 1.7578125   3
7289973 1.85546875  3

5 answers

solution1
 2  2019-05-02 13:01:36

One way would be to create a dictionary to replace the values in col_2 . In order to replace only those that appear after a 19 , GroupBy , check equality and take the cumsum to perform boolean indexation on the dataframe:

map_ = {1:25, 2:26, 3:27}
cs = df.col_2.eq(19).groupby(df.col_1).cumsum()
update = df.loc[cs].col_2.replace(map_)
df.loc[update.index, 'col_2'] = update

 col_1  col_2
0       1      1
1       1      2
2       1      3
3       1      6
4       1      8
5       1     11
6       1     12
7       1     19
8       1     25
9       1     26
10      1     27
11      2      1
12      2      2
13      2      4
14      2      6
15      2      7
16      2     11
17      2     13
18      2     16
19      2     19
20      2     25
21      2     26
22      2     27

solution2
 0  2019-05-02 13:08:22

My try:

def fill19(x):
    # x.shift()==19 marks all 1's after 19's
    # rolling(3) marks three numbers after 19's
    filters = (x.shift()==19).rolling(3).sum().fillna(0).astype(bool)
    x[filters] += 24
    return x

df.col2 = df.groupby('col_1').col_2.apply(fill19)

0      1
1      2
2      3
3      6
4      8
5     11
6     12
7     19
8     25
9     26
10    27
11     1
12     2
13     4
14     6
15     7
16    11
17    13
18    16
19    19
20    25
21    26
22    27
Name: col_2, dtype: int64

solution3
 0  2019-05-02 13:08:52

Try below for loop, it does what is required in your case:

for i in df['col_1'].unique():
     ix = np.argwhere((df['col_1'] == i) & (df['col_2'] == 19 ))
     df.loc[ix[0][0]+1, 'col_2'] = 25
     df.loc[ix[0][0]+2, 'col_2'] = 26
     df.loc[ix[0][0]+3, 'col_2'] = 27

solution4
 0  2019-05-02 13:47:58

Old school looping

map_ = {1: 25, 2: 26, 3: 27}

d = {}  # Tracks if 19 has been seen yet

for i, c1, c2 in df.itertuples():

    if d.setdefault(c1, False):
        df.at[i, 'col_2'] = map_.get(c2, c2)

    d[c1] |= c2 == 19

Using np.logigcal_or

m = df.col_2.eq(19)
m = m.groupby(df.col_1).transform(np.logical_or.accumulate) ^ m
df.assign(col_2=df.col_2 + m * 24)

    col_1  col_2
0       1      1
1       1      2
2       1      3
3       1      6
4       1      8
5       1     11
6       1     12
7       1     19
8       1     25
9       1     26
10      1     27
11      2      1
12      2      2
13      2      4
14      2      6
15      2      7
16      2     11
17      2     13
18      2     16
19      2     19
20      2     25
21      2     26
22      2     27

solution5
 0 ACCPTED  2019-05-02 14:27:51

Based on your real DataFrame, you can do the following:

df['StepID'] = df.groupby('ContextID')['StepID'].apply(
    lambda x: x + (x < x.shift(1)).cumsum()*24)

print(df.tail(25))

Output:

     ContextID  BacksGas_Flow_sccm  StepID
138    7289973            1.855469      24
139    7289973            1.757812      24
140    7289973            1.855469      24
141    7289973            1.855469      24
142    7289973            1.757812      24
143    7289973            1.757812      25
144    7289973            1.855469      25
145    7289973            1.855469      25
146    7289973            1.855469      26
147    7289973            1.757812      26
148    7289973            1.953125      26
149    7289973            1.953125      26
150    7289973            1.855469      26
151    7289973            1.855469      27
152    7289973            1.855469      27
153    7289973            1.855469      27
154    7289973            1.953125      27
155    7289973            1.855469      27
156    7289973            1.953125      27
157    7289973            1.855469      27
158    7289973            1.757812      27
159    7289973            1.855469      27
160    7289973            1.855469      27
161    7289973            1.757812      27
162    7289973            1.855469      27

PS In your sample data there is only one ContextID , but I'm assuming there might be others as well in the full dataset, so I've added groupby

Update : the following is if you only need to increment values after 24 for each ContextID once by 24 (I'm saving new values to StepID_new column to show before-and-after the transformation):

x = [1,2,3,19,19,24,1,1,2,2,3,3,2,3,1]
df = pd.DataFrame({'ContextID': np.repeat([1,2], len(x)),
                   'StepID': x * 2})

df['StepID_new'] = df['StepID'] + df.groupby('ContextID')['StepID'].transform(
    lambda x: ((x==24).cumsum() > 0).shift(1, fill_value=0) * 25)

print(df)

Output:

    ContextID  StepID  StepID_new
0           1       1           1
1           1       2           2
2           1       3           3
3           1      19          19
4           1      19          19
5           1      24          24
6           1       1          26
7           1       1          26
8           1       2          27
9           1       2          27
10          1       3          28
11          1       3          28
12          1       2          27
13          1       3          28
14          1       1          26
15          2       1           1
16          2       2           2
17          2       3           3
18          2      19          19
19          2      19          19
20          2      24          24
21          2       1          26
22          2       1          26
23          2       2          27
24          2       2          27
25          2       3          28
26          2       3          28
27          2       2          27
28          2       3          28
29          2       1          26

暂无
暂无

The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.

Related Question How to replace nan values of a column based on certain values of other column How to average certain values of a column based on other columns condition in pandas How to apply a certain function to DataFrame column values on a certain condition Replace values within a column if a certain condition is met using Python Replace NaN with values from other column based on certain conditions Replace certain values depending on condition in pandas Replace row values by condition if they are in certain time range How to create a column of zeros of a particular value based on a certain condition? How to select certain values based on a condition in a data frame? How do I get a series with certain values based on a condition in a dataframe?
Related Tags
 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM