如何在 pandas 中添加行直到满足条件

Question

I have a dataframe like this:

x     y    frames
0    10     7 
1    14     8    
1    19     9
3    11     10

I want to add rows to the top of the dataframe so that frames can start at 0 and increment up by 1 until it reaches the first original frames value. I want to add NaN values for all other columns.

I tried this, but it takes forever and eventually crashes:

starting_frame = int(df.frames.values[0])
print(starting_frame)
count = 0
while count < starting_frame:
    row = ['', np.nan, np.nan, np.nan,
       np.nan, np.nan, np.nan,
       '', np.nan, np.nan, np.nan,
       count]
    d = pd.DataFrame(row)
    df = pd.concat([d, df])
    count+=1

print(df)

It should look like this:

x     y    frames
NaN   NaN     0 
NaN   NaN     1 
NaN   NaN     2 
NaN   NaN     3 
NaN   NaN     4 
NaN   NaN     5 
NaN   NaN     6 
0      10     7 
1      14     8    
1      19     9
3      11     10

Answer 1

Try with reindex

out = df.set_index('frames').reindex(range(df['frames'].max()+1)).reset_index()

out
Out[545]: 
    frames    x     y
0        0  NaN   NaN
1        1  NaN   NaN
2        2  NaN   NaN
3        3  NaN   NaN
4        4  NaN   NaN
5        5  NaN   NaN
6        6  NaN   NaN
7        7  0.0  10.0
8        8  1.0  14.0
9        9  1.0  19.0
10      10  3.0  11.0

Answer 2

You can create a Series for the missing values of frames and use concat to preprend it to you DataFrame

In [66]: pd.concat([pd.Series(range(0, df.loc[0, 'frames']), name='frames').to_frame(), df]).reset_index(drop=True)                                                                                        
Out[66]: 
    frames    x     y
0        0  NaN   NaN
1        1  NaN   NaN
2        2  NaN   NaN
3        3  NaN   NaN
4        4  NaN   NaN
5        5  NaN   NaN
6        6  NaN   NaN
7        7  0.0  10.0
8        8  1.0  14.0
9        9  1.0  19.0
10      10  3.0  11.0

Answer 3

Problems in your code:

1. Using pd.DataFrame(row) , you are creating a series rather than a row in dataframe.
2. Using pd.concat([d, df]) and count += 1 , the bigger frames will be on top.

To fix these, you could use

starting_frame = int(df.frames.values[0])

count = starting_frame - 1

while count >= 0:
    row = [np.nan, np.nan, count]

    d = pd.DataFrame([row], columns=['x', 'y', 'frames'])

    df = pd.concat([d, df])
    count -= 1

print(df)

You can also create a helper dataframe whose frames column value starts from 0, and merge df on that helper dataframe.

df = df.merge(pd.DataFrame({'frames': range(df['frames'].max() + 1)}), how='right')

# print(df)

      x     y  frames
0   NaN   NaN       0
1   NaN   NaN       1
2   NaN   NaN       2
3   NaN   NaN       3
4   NaN   NaN       4
5   NaN   NaN       5
6   NaN   NaN       6
7   0.0  10.0       7
8   1.0  14.0       8
9   1.0  19.0       9
10  3.0  11.0      10