Ejemplo Práctico 01¶

Descripción del conjunto de datos¶

Introducción¶

A continuación analizaremos el conjunto de datos denominado player_data.csv, el cual contiene información básica de los jugadores de la NBA.

Descripción de las Columnas¶

| columna | tipo | descripción | |------------ |--------- |-------------------------------------- | | name | object | nombre del jugador | | year_start | int64 | año de inicio de carrera en la NBA | | year_end | int64 | año de término de carrera en la NBA | | position | object | posición del jugador | | height | object | altura del jugador | | weight | float64 | peso del jugador | | birth_date | object | fecha de nacimiento | | college | object | universidad antes de entrar a la NBA |

Resolución del Problema¶

In [1]:

# load data
import pandas as pd
path =  "https://raw.githubusercontent.com/fralfaro/python_data_manipulation/main/docs/pandas/data/player_data.csv"
player_data = pd.read_csv(
    path, # path
    sep="," # separation
)
player_data

# load data import pandas as pd path = "https://raw.githubusercontent.com/fralfaro/python_data_manipulation/main/docs/pandas/data/player_data.csv" player_data = pd.read_csv( path, # path sep="," # separation ) player_data

Out[1]:

	name	year_start	year_end	position	height	weight	birth_date	college
0	Alaa Abdelnaby	1991	1995	F-C	6-10	240.0	June 24, 1968	Duke University
1	Zaid Abdul-Aziz	1969	1978	C-F	6-9	235.0	April 7, 1946	Iowa State University
2	Kareem Abdul-Jabbar	1970	1989	C	7-2	225.0	April 16, 1947	University of California, Los Angeles
3	Mahmoud Abdul-Rauf	1991	2001	G	6-1	162.0	March 9, 1969	Louisiana State University
4	Tariq Abdul-Wahad	1998	2003	F	6-6	223.0	November 3, 1974	San Jose State University
...	...	...	...	...	...	...	...	...
4545	Ante Zizic	2018	2018	F-C	6-11	250.0	January 4, 1997	NaN
4546	Jim Zoet	1983	1983	C	7-1	240.0	December 20, 1953	Kent State University
4547	Bill Zopf	1971	1971	G	6-1	170.0	June 7, 1948	Duquesne University
4548	Ivica Zubac	2017	2018	C	7-1	265.0	March 18, 1997	NaN
4549	Matt Zunic	1949	1949	G-F	6-3	195.0	December 19, 1919	George Washington University

4550 rows × 8 columns

Módulos Básicos¶

Existen módulos para comprender rápidamente la naturaleza del dataframe.

In [2]:

# primeras silas
print("first 5 rows:")
player_data.head(5)

# primeras silas print("first 5 rows:") player_data.head(5)

first 5 rows:

Out[2]:

	name	year_start	year_end	position	height	weight	birth_date	college
0	Alaa Abdelnaby	1991	1995	F-C	6-10	240.0	June 24, 1968	Duke University
1	Zaid Abdul-Aziz	1969	1978	C-F	6-9	235.0	April 7, 1946	Iowa State University
2	Kareem Abdul-Jabbar	1970	1989	C	7-2	225.0	April 16, 1947	University of California, Los Angeles
3	Mahmoud Abdul-Rauf	1991	2001	G	6-1	162.0	March 9, 1969	Louisiana State University
4	Tariq Abdul-Wahad	1998	2003	F	6-6	223.0	November 3, 1974	San Jose State University

In [3]:

# ultimas filas
print("\nlast 5 rows:")
player_data.tail(5)

# ultimas filas print("\nlast 5 rows:") player_data.tail(5)

last 5 rows:

Out[3]:

	name	year_start	year_end	position	height	weight	birth_date	college
4545	Ante Zizic	2018	2018	F-C	6-11	250.0	January 4, 1997	NaN
4546	Jim Zoet	1983	1983	C	7-1	240.0	December 20, 1953	Kent State University
4547	Bill Zopf	1971	1971	G	6-1	170.0	June 7, 1948	Duquesne University
4548	Ivica Zubac	2017	2018	C	7-1	265.0	March 18, 1997	NaN
4549	Matt Zunic	1949	1949	G-F	6-3	195.0	December 19, 1919	George Washington University

In [4]:

# tipo
print("\ntype of dataframe:")
type(player_data)

# tipo print("\ntype of dataframe:") type(player_data)

type of dataframe:

Out[4]:

pandas.core.frame.DataFrame

In [5]:

# tipo por columnas
print("\ntype of columns:")
player_data.dtypes

# tipo por columnas print("\ntype of columns:") player_data.dtypes

type of columns:

Out[5]:

name           object
year_start      int64
year_end        int64
position       object
height         object
weight        float64
birth_date     object
college        object
dtype: object

In [6]:

# dimension
print("\nshape:")
player_data.shape

# dimension print("\nshape:") player_data.shape

shape:

Out[6]:

(4550, 8)

In [7]:

# nombre de las columnas
print("\ncols:")
player_data.columns

# nombre de las columnas print("\ncols:") player_data.columns

cols:

Out[7]:

Index(['name', 'year_start', 'year_end', 'position', 'height', 'weight',
       'birth_date', 'college'],
      dtype='object')

In [8]:

# indice
print("\nindex:")
player_data.index

# indice print("\nindex:") player_data.index

index:

Out[8]:

RangeIndex(start=0, stop=4550, step=1)

In [9]:

# acceder a la columna posicion
print("\ncolumn 'position': ")
player_data['position'].head()

# acceder a la columna posicion print("\ncolumn 'position': ") player_data['position'].head()

column 'position': 

Out[9]:

0    F-C
1    C-F
2      C
3      G
4      F
Name: position, dtype: object

In [10]:

# cambiar nombre de una o varias columnas
player_data = player_data.rename(columns={"birth_date": "Birth", "college": "College"})
player_data.head()

# cambiar nombre de una o varias columnas player_data = player_data.rename(columns={"birth_date": "Birth", "college": "College"}) player_data.head()

Out[10]:

	name	year_start	year_end	position	height	weight	Birth	College
0	Alaa Abdelnaby	1991	1995	F-C	6-10	240.0	June 24, 1968	Duke University
1	Zaid Abdul-Aziz	1969	1978	C-F	6-9	235.0	April 7, 1946	Iowa State University
2	Kareem Abdul-Jabbar	1970	1989	C	7-2	225.0	April 16, 1947	University of California, Los Angeles
3	Mahmoud Abdul-Rauf	1991	2001	G	6-1	162.0	March 9, 1969	Louisiana State University
4	Tariq Abdul-Wahad	1998	2003	F	6-6	223.0	November 3, 1974	San Jose State University

In [11]:

# fijar columna especifica como indice
player_data = player_data.set_index(["name"])
player_data.head()

# fijar columna especifica como indice player_data = player_data.set_index(["name"]) player_data.head()

Out[11]:

	year_start	year_end	position	height	weight	Birth	College
name
Alaa Abdelnaby	1991	1995	F-C	6-10	240.0	June 24, 1968	Duke University
Zaid Abdul-Aziz	1969	1978	C-F	6-9	235.0	April 7, 1946	Iowa State University
Kareem Abdul-Jabbar	1970	1989	C	7-2	225.0	April 16, 1947	University of California, Los Angeles
Mahmoud Abdul-Rauf	1991	2001	G	6-1	162.0	March 9, 1969	Louisiana State University
Tariq Abdul-Wahad	1998	2003	F	6-6	223.0	November 3, 1974	San Jose State University

In [12]:

# ordenar dataframe por columna especifica
player_data = player_data.sort_values("weight")
player_data.head()

# ordenar dataframe por columna especifica player_data = player_data.sort_values("weight") player_data.head()

Out[12]:

	year_start	year_end	position	height	weight	Birth	College
name
Penny Early	1969	1969	G	5-3	114.0	May 30, 1943	NaN
Spud Webb	1986	1998	G	5-6	133.0	July 13, 1963	North Carolina State University
Earl Boykins	1999	2012	G	5-5	135.0	June 2, 1976	Eastern Michigan University
Muggsy Bogues	1988	2001	G	5-3	136.0	January 9, 1965	Wake Forest University
Chet Aubuchon	1947	1947	G	5-10	137.0	May 18, 1916	Michigan State University

In [13]:

# resumen de la información 
player_data.describe(include='all')# player_data.describe()

# resumen de la información player_data.describe(include='all')# player_data.describe()

Out[13]:

	year_start	year_end	position	height	weight	Birth	College
count	4550.000000	4550.000000	4549	4549	4544.000000	4519	4248
unique	NaN	NaN	7	28	NaN	4161	473
top	NaN	NaN	G	6-7	NaN	February 23, 1945	University of Kentucky
freq	NaN	NaN	1574	473	NaN	3	99
mean	1985.076264	1989.272527	NaN	NaN	208.908011	NaN	NaN
std	20.974188	21.874761	NaN	NaN	26.268662	NaN	NaN
min	1947.000000	1947.000000	NaN	NaN	114.000000	NaN	NaN
25%	1969.000000	1973.000000	NaN	NaN	190.000000	NaN	NaN
50%	1986.000000	1992.000000	NaN	NaN	210.000000	NaN	NaN
75%	2003.000000	2009.000000	NaN	NaN	225.000000	NaN	NaN
max	2018.000000	2018.000000	NaN	NaN	360.000000	NaN	NaN

Módulos Avanzados¶

Cuando se trabaja con un conjunto de datos, se crea una dinámica de preguntas y respuestas, en donde a medida que necesito información, se va accediendo al dataframe. En algunas ocaciones es directo, basta un simple módulo, aunque en otras será necesaria realizar operaciones un poco más complejas.

Por ejemplo, del conjunto de datos en estudio, se esta interesado en responder las siguientes preguntas:

a) Determine si el dataframe tiene valores nulos

In [14]:

# veamos las columnas con datos nulos
for col in player_data.columns:
   
    temp = player_data[player_data[col].isnull()]
    print(col)
    print(f"valores nulos: {len(temp)}\n")

# veamos las columnas con datos nulos for col in player_data.columns: temp = player_data[player_data[col].isnull()] print(col) print(f"valores nulos: {len(temp)}\n")

year_start
valores nulos: 0

year_end
valores nulos: 0

position
valores nulos: 1

height
valores nulos: 1

weight
valores nulos: 6

Birth
valores nulos: 31

College
valores nulos: 302

In [15]:

# ocupar comando .notnull().all(axis=1) para ver todos los valores que NO son nulos para todas las columnas
player_data.notnull().all(axis=1).head(10)

# ocupar comando .notnull().all(axis=1) para ver todos los valores que NO son nulos para todas las columnas player_data.notnull().all(axis=1).head(10)

Out[15]:

name
Penny Early        False
Spud Webb           True
Earl Boykins        True
Muggsy Bogues       True
Chet Aubuchon       True
Greg Grant          True
Angelo Musi         True
Ernie Calverley     True
Tyler Ulis          True
Lionel Malamed      True
dtype: bool

b) Elimine los valores nulos del dataframe

In [16]:

# ocupar masking
mask = lambda df: df.notnull().all(axis=1)
player_data = player_data[mask]
player_data.head()

# ocupar masking mask = lambda df: df.notnull().all(axis=1) player_data = player_data[mask] player_data.head()

Out[16]:

	year_start	year_end	position	height	weight	Birth	College
name
Spud Webb	1986	1998	G	5-6	133.0	July 13, 1963	North Carolina State University
Earl Boykins	1999	2012	G	5-5	135.0	June 2, 1976	Eastern Michigan University
Muggsy Bogues	1988	2001	G	5-3	136.0	January 9, 1965	Wake Forest University
Chet Aubuchon	1947	1947	G	5-10	137.0	May 18, 1916	Michigan State University
Greg Grant	1990	1996	G	5-7	140.0	August 29, 1966	Trenton State University

c) Determinar el tiempo (en años) de cada jugador en su posición

In [17]:

player_data['duration'] = player_data['year_end'] - player_data['year_start']
player_data.head()

player_data['duration'] = player_data['year_end'] - player_data['year_start'] player_data.head()

Out[17]:

	year_start	year_end	position	height	weight	Birth	College	duration
name
Spud Webb	1986	1998	G	5-6	133.0	July 13, 1963	North Carolina State University	12
Earl Boykins	1999	2012	G	5-5	135.0	June 2, 1976	Eastern Michigan University	13
Muggsy Bogues	1988	2001	G	5-3	136.0	January 9, 1965	Wake Forest University	13
Chet Aubuchon	1947	1947	G	5-10	137.0	May 18, 1916	Michigan State University	0
Greg Grant	1990	1996	G	5-7	140.0	August 29, 1966	Trenton State University	6

d) Fecha de str a objeto datetime

In [18]:

player_data['birth_date_dt'] = pd.to_datetime(player_data['Birth'], format="%B %d, %Y")
player_data.head()

player_data['birth_date_dt'] = pd.to_datetime(player_data['Birth'], format="%B %d, %Y") player_data.head()

Out[18]:

	year_start	year_end	position	height	weight	Birth	College	duration	birth_date_dt
name
Spud Webb	1986	1998	G	5-6	133.0	July 13, 1963	North Carolina State University	12	1963-07-13
Earl Boykins	1999	2012	G	5-5	135.0	June 2, 1976	Eastern Michigan University	13	1976-06-02
Muggsy Bogues	1988	2001	G	5-3	136.0	January 9, 1965	Wake Forest University	13	1965-01-09
Chet Aubuchon	1947	1947	G	5-10	137.0	May 18, 1916	Michigan State University	0	1916-05-18
Greg Grant	1990	1996	G	5-7	140.0	August 29, 1966	Trenton State University	6	1966-08-29

e) Determinar todas las posiciones

In [19]:

positions = player_data['position'].unique()
positions

positions = player_data['position'].unique() positions

Out[19]:

array(['G', 'G-F', 'F-G', 'F', 'F-C', 'C-F', 'C'], dtype=object)

f) Iterar sobre cada posición y encontrar el mayor valor de la columna weight

In [20]:

# Iterar sobre cada posición y encontrar el mayor valor.
nba_position_duration = dict()

# iterar
for position in positions:
    
    # filtrar 
    df_aux = player_data.loc[lambda x: x['position'] == position]
    
    # encontrar maximo de la columna objetivo
    max_duration = df_aux['weight'].max()
    
    # guardar en pd.Series
    nba_position_duration[position] = max_duration
    
# retornar serie
nba_position_duration

# Iterar sobre cada posición y encontrar el mayor valor. nba_position_duration = dict() # iterar for position in positions: # filtrar df_aux = player_data.loc[lambda x: x['position'] == position] # encontrar maximo de la columna objetivo max_duration = df_aux['weight'].max() # guardar en pd.Series nba_position_duration[position] = max_duration # retornar serie nba_position_duration

Out[20]:

{'G': 235.0,
 'G-F': 240.0,
 'F-G': 245.0,
 'F': 284.0,
 'F-C': 290.0,
 'C-F': 280.0,
 'C': 360.0}