Apply each value in a column against the whole column

pip install a-pandas-ex-apply-against-all

from a_pandas_ex_apply_against_all import pd_add_apply_each

import pandas as pd

pd_add_apply_each()

df = pd.read_csv(

    "https://raw.githubusercontent.com/pandas-dev/pandas/main/doc/data/titanic.csv"

)

df1 = df.PassengerId.s_apply_each(

    expression="str(x) + str(y)", # use always x/y in your expression

    exception_value=pd.NA,

    diagonal_value=pd.NA,

    print_exception=True,

)

print(df1)

      0     1     2     3     4    ...     886     887     888     889     890

0    <NA>    12    13    14    15  ...    1887    1888    1889    1890    1891

1      21  <NA>    23    24    25  ...    2887    2888    2889    2890    2891

2      31    32  <NA>    34    35  ...    3887    3888    3889    3890    3891

3      41    42    43  <NA>    45  ...    4887    4888    4889    4890    4891

4      51    52    53    54  <NA>  ...    5887    5888    5889    5890    5891

..    ...   ...   ...   ...   ...  ...     ...     ...     ...     ...     ...

886  8871  8872  8873  8874  8875  ...    <NA>  887888  887889  887890  887891

887  8881  8882  8883  8884  8885  ...  888887    <NA>  888889  888890  888891

888  8891  8892  8893  8894  8895  ...  889887  889888    <NA>  889890  889891

889  8901  8902  8903  8904  8905  ...  890887  890888  890889    <NA>  890891

890  8911  8912  8913  8914  8915  ...  891887  891888  891889  891890    <NA>

[891 rows x 891 columns]



# If you use a non-built-in function, you have to pass it as an argument, and use it as "func" in your expression

# An example using shapely (merging different polygons)



from shapely.ops import unary_union

import shapely

polyshape = []

for k in range(10):

    xmin = k * 10 + 5

    ymin = k * 10 + 5

    xmax = k * 20 + 10

    ymax = k * 20 + 10

    coordsalls = [[xmin, ymin], [xmax, ymin], [xmax, ymax], [xmin, ymax], [xmin, ymin]]

    po = shapely.geometry.Polygon(coordsalls)

    polyshape.append(po)

df2 = pd.DataFrame(polyshape)



df2

Out[7]: 

                                                   0

0            POLYGON ((5 5, 10 5, 10 10, 5 10, 5 5))

1      POLYGON ((15 15, 30 15, 30 30, 15 30, 15 15))

2      POLYGON ((25 25, 50 25, 50 50, 25 50, 25 25))

3      POLYGON ((35 35, 70 35, 70 70, 35 70, 35 35))

4      POLYGON ((45 45, 90 45, 90 90, 45 90, 45 45))

5  POLYGON ((55 55, 110 55, 110 110, 55 110, 55 55))

6  POLYGON ((65 65, 130 65, 130 130, 65 130, 65 65))

7  POLYGON ((75 75, 150 75, 150 150, 75 150, 75 75))

8  POLYGON ((85 85, 170 85, 170 170, 85 170, 85 85))

9  POLYGON ((95 95, 190 95, 190 190, 95 190, 95 95))





df1 = df2[0].s_apply_each(

    expression="func([x,y]) if x.intersects(y) else x", # use always x/y in your expression

    func=unary_union,

    exception_value=pd.NA,

    diagonal_value=pd.NA,

    print_exception=True,

    ignore_exceptions=True

)



Out[8]: 

                                                   0  ...                                                  9

0                                               <NA>  ...            POLYGON ((5 5, 10 5, 10 10, 5 10, 5 5))

1      POLYGON ((15 15, 30 15, 30 30, 15 30, 15 15))  ...      POLYGON ((15 15, 30 15, 30 30, 15 30, 15 15))

2      POLYGON ((25 25, 50 25, 50 50, 25 50, 25 25))  ...      POLYGON ((25 25, 50 25, 50 50, 25 50, 25 25))

3      POLYGON ((35 35, 70 35, 70 70, 35 70, 35 35))  ...      POLYGON ((35 35, 70 35, 70 70, 35 70, 35 35))

4      POLYGON ((45 45, 90 45, 90 90, 45 90, 45 45))  ...      POLYGON ((45 45, 90 45, 90 90, 45 90, 45 45))

5  POLYGON ((55 55, 110 55, 110 110, 55 110, 55 55))  ...  POLYGON ((110 55, 55 55, 55 110, 95 110, 95 19...

6  POLYGON ((65 65, 130 65, 130 130, 65 130, 65 65))  ...  POLYGON ((130 65, 65 65, 65 130, 95 130, 95 19...

7  POLYGON ((75 75, 150 75, 150 150, 75 150, 75 75))  ...  POLYGON ((150 75, 75 75, 75 150, 95 150, 95 19...

8  POLYGON ((85 85, 170 85, 170 170, 85 170, 85 85))  ...  POLYGON ((170 85, 85 85, 85 170, 95 170, 95 19...

9  POLYGON ((95 95, 190 95, 190 190, 95 190, 95 95))  ...                                               <NA>

[10 rows x 10 columns]