mutate

In [1]:

# https://dplyr.tidyverse.org/reference/mutate.html
%run nb_helpers.py

from datar.data import starwars
from datar.all import *

nb_header(mutate, transmute)

# https://dplyr.tidyverse.org/reference/mutate.html %run nb_helpers.py from datar.data import starwars from datar.all import * nb_header(mutate, transmute)

Try this notebook on binder.

★ mutate

¶

Add new columns to a data frame.¶

The original API:
https://dplyr.tidyverse.org/reference/mutate.html

Args:¶

  _data: A data frame
  _keep: allows you to control which columns from _data are retained
    in the output:

    - "all", the default, retains all variables.

    - "used" keeps any variables used to make new variables;       it's useful for checking your work as it displays inputs and
      outputs side-by-side.

    - "unused" keeps only existing variables not used to make new       variables.

    - "none", only keeps grouping keys (like transmute()).

  _before: A list of column names to put the new columns before.
  _after: A list of column names to put the new columns after.
  *args: and
  **kwargs: Name-value pairs. The name gives the name of the column
    in the output. The value can be:

    - A vector of length 1, which will be recycled to the correct       length.

    - A vector the same length as the current group (or the whole       data frame if ungrouped).

    - None to remove the column

Returns:¶

  An object of the same type as _data. The output has the following
  properties:
  - Rows are not affected.   - Existing columns will be preserved according to the _keep     argument. New columns will be placed according to the
    _before and _after arguments. If _keep = "none"
    (as in transmute()), the output order is determined only
    by ..., not the order of existing columns.

  - Columns given value None will be removed   - Groups will be recomputed if a grouping variable is mutated.   - Data frame attributes are preserved.

★ transmute

¶

Add new columns to a data frame and remove existing columns¶

using mutate with _keep="none".

The original API:
https://dplyr.tidyverse.org/reference/mutate.html

Args:¶

  _data: A data frame
  _before: A list of column names to put the new columns before.
  _after: A list of column names to put the new columns after.
  *args: and
  **kwargs: Name-value pairs. The name gives the name of the column
    in the output. The value can be:

    - A vector of length 1, which will be recycled to the correct       length.

    - A vector the same length as the current group (or the whole       data frame if ungrouped).

    - None to remove the column

Returns:¶

  An object of the same type as _data. The output has the following
  properties:
  - Rows are not affected.   - Existing columns will be preserved according to the _keep     argument. New columns will be placed according to the
    _before and _after arguments. If _keep = "none"
    (as in transmute()), the output order is determined only
    by ..., not the order of existing columns.

  - Columns given value None will be removed   - Groups will be recomputed if a grouping variable is mutated.   - Data frame attributes are preserved.

In [2]:

starwars >> \
 select(f.name, f.mass) >> \
 mutate(
  mass2 = f.mass * 2,
  mass2_squared = f.mass2 * f.mass2
)

starwars >> \ select(f.name, f.mass) >> \ mutate( mass2 = f.mass * 2, mass2_squared = f.mass2 * f.mass2 )

Out[2]:

	name	mass	mass2	mass2_squared
	<object>	<float64>	<float64>	<float64>
0	Luke Skywalker	77.0	154.0	23716.0
1	C-3PO	75.0	150.0	22500.0
2	R2-D2	32.0	64.0	4096.0
3	Darth Vader	136.0	272.0	73984.0
...	...	...	...	...
4	Leia Organa	49.0	98.0	9604.0
82	Rey	NaN	NaN	NaN
83	Poe Dameron	NaN	NaN	NaN
84	BB8	NaN	NaN	NaN
85	Captain Phasma	NaN	NaN	NaN
86	Padmé Amidala	45.0	90.0	8100.0

87 rows × 4 columns

In [3]:

starwars >> \
 select(f.name, f.height, f.mass, f.homeworld) >> \
 mutate(
  mass = NULL,
  height = f.height * 0.0328084 # convert to feet
)

starwars >> \ select(f.name, f.height, f.mass, f.homeworld) >> \ mutate( mass = NULL, height = f.height * 0.0328084 # convert to feet )

Out[3]:

	name	height	homeworld
	<object>	<float64>	<object>
0	Luke Skywalker	5.643045	Tatooine
1	C-3PO	5.479003	Tatooine
2	R2-D2	3.149606	Naboo
3	Darth Vader	6.627297	Tatooine
...	...	...	...
4	Leia Organa	4.921260	Alderaan
82	Rey	NaN	NaN
83	Poe Dameron	NaN	NaN
84	BB8	NaN	NaN
85	Captain Phasma	NaN	NaN
86	Padmé Amidala	5.413386	Naboo

87 rows × 3 columns

In [5]:

x = starwars >> \
 select(f.name, f.homeworld, f.species) >> \
 mutate(across(~f.name, as_factor))

x.dtypes
x

x = starwars >> \ select(f.name, f.homeworld, f.species) >> \ mutate(across(~f.name, as_factor)) x.dtypes x

Out[5]:

name           object
homeworld    category
species      category
dtype: object

Out[5]:

	name	homeworld	species
	<object>	<category>	<category>
0	Luke Skywalker	Tatooine	Human
1	C-3PO	Tatooine	Droid
2	R2-D2	Naboo	Droid
3	Darth Vader	Tatooine	Human
...	...	...	...
4	Leia Organa	Alderaan	Human
82	Rey	NaN	Human
83	Poe Dameron	NaN	Human
84	BB8	NaN	Droid
85	Captain Phasma	NaN	NaN
86	Padmé Amidala	Naboo	Human

87 rows × 3 columns

In [6]:

starwars >> \
 select(f.name, f.mass, f.homeworld) >> \
 group_by(f.homeworld) >> \
 mutate(rank=min_rank(desc(f.mass)))

starwars >> \ select(f.name, f.mass, f.homeworld) >> \ group_by(f.homeworld) >> \ mutate(rank=min_rank(desc(f.mass)))

Out[6]:

	name	mass	homeworld	rank
	<object>	<float64>	<object>	<float64>
0	Luke Skywalker	77.0	Tatooine	5.0
1	C-3PO	75.0	Tatooine	6.0
2	R2-D2	32.0	Naboo	6.0
3	Darth Vader	136.0	Tatooine	1.0
...	...	...	...	...
4	Leia Organa	49.0	Alderaan	2.0
82	Rey	NaN	NaN	NaN
83	Poe Dameron	NaN	NaN	NaN
84	BB8	NaN	NaN	NaN
85	Captain Phasma	NaN	NaN	NaN
86	Padmé Amidala	45.0	Naboo	5.0

87 rows × 4 columns

TibbleGrouped: homeworld (n=49)

In [7]:

df = tibble(x=1, y=2)

df >> mutate(z=f.x+f.y)

df = tibble(x=1, y=2) df >> mutate(z=f.x+f.y)

Out[7]:

	x	y	z
	<int64>	<int64>	<int64>
0	1	2	3

In [8]:

df >> mutate(z=f.x+f.y, _before=1)

df >> mutate(z=f.x+f.y, _before=1)

Out[8]:

	x	z	y
	<int64>	<int64>	<int64>
0	1	3	2

In [9]:

df >> mutate(z=f.x+f.y, _after=f.x)

df >> mutate(z=f.x+f.y, _after=f.x)

Out[9]:

	x	z	y
	<int64>	<int64>	<int64>
0	1	3	2

In [10]:

# use a temporary column
df >> mutate(_z=f.x+f.y, z=f._z*2, _after=f.x)

# use a temporary column df >> mutate(_z=f.x+f.y, z=f._z*2, _after=f.x)

Out[10]:

	x	z	y
	<int64>	<int64>	<int64>
0	1	6	2

In [11]:

df = tibble(x=1, y=2, a="a", b="b")
df >> mutate(z=f.x+f.y, _keep='all')

df = tibble(x=1, y=2, a="a", b="b") df >> mutate(z=f.x+f.y, _keep='all')

Out[11]:

	x	y	a	b	z
	<int64>	<int64>	<object>	<object>	<int64>
0	1	2	a	b	3

In [12]:

df >> mutate(z=f.x+f.y, _keep='used')

df >> mutate(z=f.x+f.y, _keep='used')

Out[12]:

	x	y	z
	<int64>	<int64>	<int64>
0	1	2	3

In [13]:

df >> mutate(z=f.x+f.y, _keep='unused')

df >> mutate(z=f.x+f.y, _keep='unused')

Out[13]:

	a	b	z
	<object>	<object>	<int64>
0	a	b	3

In [14]:

df >> mutate(z=f.x+f.y, _keep='none')

df >> mutate(z=f.x+f.y, _keep='none')

Out[14]:

	z
	<int64>
0	3

In [16]:

starwars >> \
  select(f.name, f.mass, f.species) >> \
  mutate(mass_norm=f.mass/mean(f.mass))

starwars >> \ select(f.name, f.mass, f.species) >> \ mutate(mass_norm=f.mass/mean(f.mass))

Out[16]:

	name	mass	species	mass_norm
	<object>	<float64>	<object>	<float64>
0	Luke Skywalker	77.0	Human	0.791270
1	C-3PO	75.0	Droid	0.770718
2	R2-D2	32.0	Droid	0.328840
3	Darth Vader	136.0	Human	1.397569
...	...	...	...	...
4	Leia Organa	49.0	Human	0.503536
82	Rey	NaN	Human	NaN
83	Poe Dameron	NaN	Human	NaN
84	BB8	NaN	Droid	NaN
85	Captain Phasma	NaN	NaN	NaN
86	Padmé Amidala	45.0	Human	0.462431

87 rows × 4 columns

In [18]:

starwars >> \
  select(f.name, f.mass, f.species) >> \
  group_by(f.species) >> \
  mutate(mass_norm=f.mass / mean(f.mass)) >> \
  ungroup()

starwars >> \ select(f.name, f.mass, f.species) >> \ group_by(f.species) >> \ mutate(mass_norm=f.mass / mean(f.mass)) >> \ ungroup()

Out[18]:

	name	mass	species	mass_norm
	<object>	<float64>	<object>	<float64>
0	Luke Skywalker	77.0	Human	0.930156
1	C-3PO	75.0	Droid	1.075269
2	R2-D2	32.0	Droid	0.458781
3	Darth Vader	136.0	Human	1.642873
...	...	...	...	...
4	Leia Organa	49.0	Human	0.591917
82	Rey	NaN	Human	NaN
83	Poe Dameron	NaN	Human	NaN
84	BB8	NaN	Droid	NaN
85	Captain Phasma	NaN	NaN	NaN
86	Padmé Amidala	45.0	Human	0.543598

87 rows × 4 columns

In [19]:

vars = ["mass", "height"]
starwars >> mutate(starwars, prod=f[vars[0]] * f[vars[1]])

vars = ["mass", "height"] starwars >> mutate(starwars, prod=f[vars[0]] * f[vars[1]])

Out[19]:

	name	height	mass	hair_color	skin_color	eye_color	birth_year	sex	gender	homeworld	species	prod
	<object>	<float64>	<float64>	<object>	<object>	<object>	<float64>	<object>	<object>	<object>	<object>	<float64>
0	Luke Skywalker	172.0	77.0	blond	fair	blue	19.0	male	masculine	Tatooine	Human	13244.0
1	C-3PO	167.0	75.0	NaN	gold	yellow	112.0	none	masculine	Tatooine	Droid	12525.0
2	R2-D2	96.0	32.0	NaN	white, blue	red	33.0	none	masculine	Naboo	Droid	3072.0
3	Darth Vader	202.0	136.0	none	white	yellow	41.9	male	masculine	Tatooine	Human	27472.0
...	...	...	...	...	...	...	...	...	...	...	...	...
4	Leia Organa	150.0	49.0	brown	light	brown	19.0	female	feminine	Alderaan	Human	7350.0
82	Rey	NaN	NaN	brown	light	hazel	NaN	female	feminine	NaN	Human	NaN
83	Poe Dameron	NaN	NaN	brown	light	brown	NaN	male	masculine	NaN	Human	NaN
84	BB8	NaN	NaN	none	none	black	NaN	none	masculine	NaN	Droid	NaN
85	Captain Phasma	NaN	NaN	unknown	unknown	unknown	NaN	NaN	NaN	NaN	NaN	NaN
86	Padmé Amidala	165.0	45.0	brown	light	brown	46.0	female	feminine	Naboo	Human	7425.0

87 rows × 12 columns