对所有正值和所有负值求和 Pandas

Question

I have the following dataframe (df):

 Row Number                                                                                       
Row 0               0.24             0.16                 -0.18           -0.20            1.24
Row 1               0.18             0.12                 -0.73           -0.36           -0.54
Row 2              -0.01             0.25                 -0.35           -0.08           -0.43
Row 3              -0.43             0.21                  0.53            0.55           -1.03
Row 4              -0.24            -0.20                  0.49            0.08            0.61
Row 5              -0.19            -0.29                 -0.08           -0.16            0.34

I am attempting to sum all the negative and positive numbers respectively, eg sum(neg_numbers) = n and sum(pos_numbers) = x

I have tried:

df.groupby(df.agg([('negative' , lambda x : x[x < 0].sum()) , ('positive' , lambda x : x[x > 0].sum())])

to no avail.

How would I sum these values?

Thank you in advance!

Answer 1

You can do

sum_pos = df[df>0].sum(1)
sum_neg = df[df<0].sum(1)

if you want to get the sums per row. If you want to sum all values regardless of rows/columns, can use np.nansum

sum_pos = np.nansum(df[df>0])

Answer 2

You can do with

df.mul(df.gt(0)).sum().sum()
Out[447]: 5.0

df.mul(~df.gt(0)).sum().sum()
Out[448]: -5.5

If need sum by row

df.mul(df.gt(0)).sum()
Out[449]: 
1    0.42
2    0.74
3    1.02
4    0.63
5    2.19
dtype: float64

Answer 3

Yet another way for the total sums:

sum_pos = df.to_numpy().flatten().clip(min=0).sum()
sum_neg = df.to_numpy().flatten().clip(max=0).sum()

And for sums by columns:

sum_pos_col = sum(df.to_numpy().clip(min=0))
sum_neg_col = sum(df.to_numpy().clip(max=0))

Answer 4

If you have string columns in dataframe and want to get the sum for particular column, then

df[df['column_name']>0]['column_name'].sum()
df[df['column_name']<0]['column_name'].sum()