It is based on the Thomas Augspurger comparison Notebook (refreshed for Pandas 0.16.0)
We just play the "R" code at the same time, instead of keeping it in comments
In [1]:
!echo %R_HOME%
In [2]:
# Some prep work to get the data from R and into pandas
%matplotlib inline
#bad tetst : move magic befor module imprort
#%load_ext rpy2.ipython
import rpy2
%load_ext rpy2.ipython
from rpy2.robjects.conversion import ri2py
from rpy2.ipython.rmagic import ri2ipython
ri2ipython.register(rpy2.robjects.Sexp, ri2py)
import numpy as np
import pandas as pd
import seaborn as sns
pd.set_option("display.max_rows", 5)
%R install.packages("tidyr")
%R install.packages("dplyr")
%R install.packages("ggplot2")
%R install.packages("rvest")
%R install.packages('RSQLite')
%R install.packages("zoo")
%R install.packages("forecast")
%R install.packages('R.utils')
%R install.packages("nycflights13")
%R install.packages('hflights')
This notebook compares pandas and dplyr. The comparison is just on syntax (verbage), not performance. Whether you're an R user looking to switch to pandas (or the other way around), I hope this guide will help ease the transition.
We'll work through the introductory dplyr vignette to analyze some flight data.
I'm working on a better layout to show the two packages side by side.
But for now I'm just putting the dplyr code in a comment above each python call.
In [3]:
%%R
library("dplyr") # for functions
library("nycflights13")
write.csv(flights, "flights.csv")
In [4]:
flights = pd.read_csv("flights.csv", index_col=0)
In [5]:
%R dim(flights)
Out[5]:
In [6]:
# dim(flights) <--- The R code
flights.shape # <--- The python code
Out[6]:
In [7]:
%R head(flights)
Out[7]:
In [8]:
# head(flights)
flights.head()
Out[8]:
dplyr has a small set of nicely defined verbs. I've listed their closest pandas verbs.
| dplyr | pandas |
| filter() (and slice()) | query() (and loc[], iloc[]) |
| arrange() | sort() |
| select() (and rename()) | \_\_getitem\_\_ (and rename()) |
| distinct() | drop_duplicates() |
| mutate() (and transmute()) | None |
| summarise() | None |
| sample_n() and sample_frac() | None |
Some of the "missing" verbs in pandas are because there are other, different ways of achieving the same goal. For example summarise is spread across mean, std, etc. Others, like sample_n, just haven't been implemented yet.
In [9]:
%R filter(flights, month == 1, day == 1)
Out[9]:
In [10]:
# filter(flights, month == 1, day == 1)
flights.query("month == 1 & day == 1")
Out[10]:
The more verbose version:
In [11]:
%R flights[flights$month == 1 & flights$day == 1, ]
Out[11]:
In [12]:
# flights[flights$month == 1 & flights$day == 1, ]
flights[(flights.month == 1) & (flights.day == 1)]
Out[12]:
In [13]:
%R slice(flights, 1:10)
Out[13]:
In [14]:
# slice(flights, 1:10)
flights.iloc[:9]
Out[14]:
In [15]:
%R arrange(flights, year, month, day)
Out[15]:
In [16]:
# arrange(flights, year, month, day)
flights.sort(['year', 'month', 'day'])
Out[16]:
In [17]:
%R arrange(flights, desc(arr_delay))
Out[17]:
In [18]:
# arrange(flights, desc(arr_delay))
flights.sort('arr_delay', ascending=False)
Out[18]:
In [19]:
%R select(flights, year, month, day)
Out[19]:
In [20]:
# select(flights, year, month, day)
flights[['year', 'month', 'day']]
Out[20]:
In [21]:
%R select(flights, year:day)
Out[21]:
In [22]:
# select(flights, year:day)
# No real equivalent here. Although I think this is OK.
# Typically I'll have the columns I want stored in a list
# somewhere, which can be passed right into __getitem__ ([]).
In [23]:
%%R
select(flights, -(year:day))
In [24]:
# select(flights, -(year:day))
# Again, simliar story. I would just use
# flights.drop(cols_to_drop, axis=1)
# or fligths[flights.columns.difference(pd.Index(cols_to_drop))]
# point to dplyr!
In [25]:
%R select(flights, tail_num = tailnum)
Out[25]:
In [26]:
# select(flights, tail_num = tailnum)
flights.rename(columns={'tailnum': 'tail_num'})['tail_num']
Out[26]:
But like Hadley mentions, not that useful since it only returns the one column. dplyr and pandas compare well here.
In [27]:
%R rename(flights, tail_num = tailnum)
Out[27]:
In [28]:
# rename(flights, tail_num = tailnum)
flights.rename(columns={'tailnum': 'tail_num'})
Out[28]:
Pandas is more verbose, but the the argument to columns can be any mapping. So it's often used with a function to perform a common task, say df.rename(columns=lambda x: x.replace('-', '_')) to replace any dashes with underscores. Also, rename (the pandas version) can be applied to the Index.
In [29]:
%R distinct(select(flights, tailnum))
Out[29]:
In [30]:
# distinct(select(flights, tailnum))
flights.tailnum.unique()
Out[30]:
FYI this returns a numpy array instead of a Series.
In [31]:
%R distinct(select(flights, origin, dest))
Out[31]:
In [32]:
# distinct(select(flights, origin, dest))
flights[['origin', 'dest']].drop_duplicates()
Out[32]:
OK, so dplyr wins there from a consistency point of view. unique is only defined on Series, not DataFrames. The original intention for drop_duplicates is to check for records that were accidentally included twice. This feels a bit hacky using it to select the distinct combinations, but it works!
In [33]:
%R mutate(flights, gain = arr_delay - dep_delay, speed = distance / air_time * 60)
Out[33]:
In [34]:
# mutate(flights,
# gain = arr_delay - dep_delay,
# speed = distance / air_time * 60)
#before pandas 0.16.0
# flights['gain'] = flights.arr_delay - flights.dep_delay
# flights['speed'] = flights.distance / flights.air_time * 60
# flights
flights.assign(gain=flights.arr_delay - flights.dep_delay,
speed=flights.distance / flights.air_time * 60)
Out[34]:
In [35]:
%R mutate(flights, gain = arr_delay - dep_delay, gain_per_hour = gain / (air_time / 60) )
Out[35]:
In [36]:
# mutate(flights,
# gain = arr_delay - dep_delay,
# gain_per_hour = gain / (air_time / 60)
# )
#before pandas 0.16.0
# flights['gain'] = flights.arr_delay - flights.dep_delay
# flights['gain_per_hour'] = flights.gain / (flights.air_time / 60)
# flights
(flights.assign(gain=flights.arr_delay - flights.dep_delay)
.assign(gain_per_hour = lambda df: df.gain / (df.air_time / 60)))
Out[36]:
The first example is pretty much identical (aside from the names, mutate vs. assign).
The second example just comes down to language differences. In R, it's possible to implement a function like mutate where you can refer to gain in the line calcuating gain_per_hour, even though gain hasn't actually been calcuated yet.
In Python, you can have arbitrary keyword arguments to functions (which we needed for .assign), but the order of the argumnets is arbitrary. So you can't have something like df.assign(x=df.a / df.b, y=x **2), because you don't know whether x or y will come first (you'd also get an error saying x is undefined.
To work around that with pandas, you'll need to split up the assigns, and pass in a callable to the second assign. The callable looks at itself to find a column named gain. Since the line above returns a DataFrame with the gain column added, the pipeline goes through just fine.
In [37]:
%R transmute(flights, gain = arr_delay - dep_delay, gain_per_hour = gain / (air_time / 60) )
Out[37]:
In [38]:
# transmute(flights,
# gain = arr_delay - dep_delay,
# gain_per_hour = gain / (air_time / 60)
# )
#before pandas 0.16.0
# flights['gain'] = flights.arr_delay - flights.dep_delay
# flights['gain_per_hour'] = flights.gain / (flights.air_time / 60)
# flights[['gain', 'gain_per_hour']]
(flights.assign(gain=flights.arr_delay - flights.dep_delay)
.assign(gain_per_hour = lambda df: df.gain / (df.air_time / 60))
[['gain', 'gain_per_hour']])
Out[38]:
In [39]:
flights.dep_delay.mean()
Out[39]:
There's an open PR on Github to make this nicer (closer to dplyr). For now you can drop down to numpy.
In [40]:
%R sample_n(flights, 10)
Out[40]:
In [41]:
# sample_n(flights, 10)
flights.loc[np.random.choice(flights.index, 10)]
Out[41]:
In [42]:
%R sample_frac(flights, 0.01)
Out[42]:
In [43]:
# sample_frac(flights, 0.01)
flights.iloc[np.random.randint(0, len(flights),
.1 * len(flights))]
Out[43]:
In [44]:
%R planes <- group_by(flights, tailnum)
%R delay <- summarise(planes, count = n(),dist = mean(distance, na.rm = TRUE), delay = mean(arr_delay, na.rm = TRUE))
%R delay <- filter(delay, count > 20, dist < 2000)
Out[44]:
In [45]:
# planes <- group_by(flights, tailnum)
# delay <- summarise(planes,
# count = n(),
# dist = mean(distance, na.rm = TRUE),
# delay = mean(arr_delay, na.rm = TRUE))
# delay <- filter(delay, count > 20, dist < 2000)
planes = flights.groupby("tailnum")
delay = (planes.agg({"year": "count",
"distance": "mean",
"arr_delay": "mean"})
.rename(columns={"distance": "dist",
"arr_delay": "delay",
"year": "count"})
.query("count > 20 & dist < 2000"))
delay
Out[45]:
For me, dplyr's n() looked is a bit starge at first, but it's already growing on me.
I think pandas is more difficult for this particular example.
There isn't as natural a way to mix column-agnostic aggregations (like count) with column-specific aggregations like the other two. You end up writing could like .agg{'year': 'count'} which reads, "I want the count of year", even though you don't care about year specifically.
Additionally assigning names can't be done as cleanly in pandas; you have to just follow it up with a rename like before.
We may as well reproduce the graph. It looks like ggplots geom_smooth is some kind of lowess smoother. We can either us seaborn:
In [46]:
import matplotlib.pyplot as plt
fig, ax = plt.subplots(figsize=(12, 6))
sns.regplot("dist", "delay", data=delay, lowess=True, ax=ax,
scatter_kws={'color': 'k', 'alpha': .5, 's': delay['count'] / 10}, ci=90,
line_kws={'linewidth': 3});
Or using statsmodels directly for more control over the lowess, with an extremely lazy "confidence interval".
In [47]:
import statsmodels.api as sm
In [48]:
smooth = sm.nonparametric.lowess(delay.delay, delay.dist, frac=1/8)
ax = delay.plot(kind='scatter', x='dist', y = 'delay', figsize=(12, 6),
color='k', alpha=.5, s=delay['count'] / 10)
ax.plot(smooth[:, 0], smooth[:, 1], linewidth=3);
std = smooth[:, 1].std()
ax.fill_between(smooth[:, 0], smooth[:, 1] - std, smooth[:, 1] + std, alpha=.25);
In [49]:
%R destinations <- group_by(flights, dest)
%R summarise(destinations, planes = n_distinct(tailnum), flights = n())
Out[49]:
In [50]:
# destinations <- group_by(flights, dest)
# summarise(destinations,
# planes = n_distinct(tailnum),
# flights = n()
# )
destinations = flights.groupby('dest')
destinations.agg({
'tailnum': lambda x: len(x.unique()),
'year': 'count'
}).rename(columns={'tailnum': 'planes',
'year': 'flights'})
Out[50]:
Similar to how dplyr provides optimized C++ versions of most of the summarise functions, pandas uses cython optimized versions for most of the agg methods.
In [51]:
%R daily <- group_by(flights, year, month, day)
%R (per_day <- summarise(daily, flights = n()))
Out[51]:
In [52]:
# daily <- group_by(flights, year, month, day)
# (per_day <- summarise(daily, flights = n()))
daily = flights.groupby(['year', 'month', 'day'])
per_day = daily['distance'].count()
per_day
Out[52]:
In [53]:
%R (per_month <- summarise(per_day, flights = sum(flights)))
Out[53]:
In [54]:
# (per_month <- summarise(per_day, flights = sum(flights)))
per_month = per_day.groupby(level=['year', 'month']).sum()
per_month
Out[54]:
In [55]:
%R (per_year <- summarise(per_month, flights = sum(flights)))
Out[55]:
In [56]:
# (per_year <- summarise(per_month, flights = sum(flights)))
per_year = per_month.sum()
per_year
Out[56]:
I'm not sure how dplyr is handling the other columns, like year, in the last example. With pandas, it's clear that we're grouping by them since they're included in the groupby. For the last example, we didn't group by anything, so they aren't included in the result.
Any follower of Hadley's twitter account will know how much R users love the %>% (pipe) operator. And for good reason!
In [57]:
%R flights %>% group_by(year, month, day) %>% select(arr_delay, dep_delay) %>% summarise( arr = mean(arr_delay, na.rm = TRUE), dep = mean(dep_delay, na.rm = TRUE)) %>% filter(arr > 30 | dep > 30)
Out[57]:
In [58]:
# flights %>%
# group_by(year, month, day) %>%
# select(arr_delay, dep_delay) %>%
# summarise(
# arr = mean(arr_delay, na.rm = TRUE),
# dep = mean(dep_delay, na.rm = TRUE)
# ) %>%
# filter(arr > 30 | dep > 30)
(
flights.groupby(['year', 'month', 'day'])
[['arr_delay', 'dep_delay']]
.mean()
.query('arr_delay > 30 | dep_delay > 30')
)
Out[58]:
Pandas has tons IO tools to help you get data in and out, including SQL databases via SQLAlchemy.
I think pandas held up pretty well, considering this was a vignette written for dplyr. I found the degree of similarity more interesting than the differences. The most difficult task was renaming of columns within an operation; they had to be followed up with a call to rename after the operation, which isn't that burdensome honestly.
More and more it looks like we're moving towards future where being a language or package partisan just doesn't make sense. Not when you can load up a Jupyter (formerly IPython) notebook to call up a library written in R, and hand those results off to python or Julia or whatever for followup, before going back to R to make a cool shiny web app.
There will always be a place for your "utility belt" package like dplyr or pandas, but it wouldn't hurt to be familiar with both.
If you want to contribute to pandas, we're always looking for help at https://github.com/pydata/pandas/. You can get ahold of me directly on twitter.
In [56]: