Reshaping Pandas Data Frame - python

I'm parsing some HTML data using Pandas like this:
rankings = pd.read_html('https://en.wikipedia.org/wiki/Rankings_of_universities_in_the_United_Kingdom')
university_guide = rankings[0]
This gives me a nice data frame:
What I want is to reshape this data frame so that there are only two columns (rank and university name). My current solution is to do something like this:
ug_copy = rankings[0][1:]
npa1 = ug_copy.as_matrix( columns=[0,1] )
npa2 = ug_copy.as_matrix( columns=[2,3] )
npa3 = ug_copy.as_matrix( columns=[4,5] )
npam = np.append(npa1,npa2)
npam = np.append(npam,npa3)
reshaped = npam.reshape((npam.size/2,2))
pd.DataFrame(data=reshaped)
This gives me what I want, but it doesn't seem like it could possibly be the best solution. I can't seem to find a good way to complete this all using a data frame. I've tried using stack/unstack and pivoting the data frame (as some of the other solutions here have suggested), but I haven't had any luck. I've tried doing something like this:
ug_copy.columns=['Rank','University','Rank','University','Rank','University']
ug_copy = ug_copy[1:]
ug_copy.groupby(['Rank', 'University'])
There has to be something small I'm missing!

This is probably a bit shorter (also note that you can use the header option in read_html to save a bit of work):
import pandas as pd
rankings = pd.read_html('https://en.wikipedia.org/wiki/Rankings_of_universities_in_the_United_Kingdom', header=0)
university_guide = rankings[0]
df = pd.DataFrame(university_guide.values.reshape((30, 2)), columns=['Rank', 'University'])
df = df.sort('Rank').reset_index(drop=True)
print df

Related

Pandas Python probelm

import pandas as pd
nba = pd.read_csv("nba.csv")
names = pd.Series(nba['Name'])
data = nba['Salary']
nba_series = (data, index=[names])
print(nba_series)
Hello I am trying to convert the columns 'Name' and 'Salary' into a series from a dataframe. I need to set the names as the index and the salaries as the values but i cannot figure it out. this is my best attempt so far anyone guidance is appreciated
I think you are over-thinking this. Simply construct it with pd.Series(). Note the data needs to be with .values, otherwis eyou'll get Nans
import pandas as pd
nba = pd.read_csv("nba.csv")
nba_series = pd.Series(data=nba['Salary'].values, index=nba['Name'])
Maybe try set_index?
nba.set_index('name', inlace = True )
nba_series = nba['Salary']
This might help you
import pandas as pd
nba = pd.read_csv("nba.csv")
names = nba['Name']
#It's automatically a series
data = nba['Salary']
#Set names as index of series
data.index = nba_series
data.index = names might be correct but depends on the data

mask function doesn't get rid of unwanted data

I'm working on a data frame taken from Adafruit IO and sadly some of my data is from a time when my project malfunctioned so some of the values are just equal NaN.
I tried to remove it by typing this code lines:
onlyValidData=temp_data.mask(temp_data['value'] =='NaN')
onlyValidData
This is data retreived from Adafruit IO Feed, getting analyzed by pandas, I tried using 'where' function too but it didn't work
my entire code is
import pandas as pd
temp_data = pd.read_json('https://io.adafruit.com/api/(...)')
light_data = pd.read_json('https://io.adafruit.com/api/(...)')
temp_data['created_at'] = pd.to_datetime(temp_data['created_at'], infer_datetime_format=True)
temp_data = temp_data.set_index('created_at')
light_data['created_at'] = pd.to_datetime(light_data['created_at'], infer_datetime_format=True)
light_data = light_data.set_index('created_at')
tempVals = pd.Series(temp_data['value'])
lightVals = pd.Series(light_data['value'])
onlyValidData=temp_data.mask(temp_data['value'] =='NaN')
onlyValidData
The output is all of my data for some reason, but it should be only the valid values.
Hey I think the issue here that you're looking for values equal to the string 'NaN', while actual NaN values aren't a string, or more specifically aren't anything.
Try using:
onlyValidData = temp_data.mask(temp_data['value'].isnull())
Edit: to remove rows rather than marking all values in that row as NaN:
onlyValidData = temp_data.dropna()

Pandas: fastest way to the DF by date

I have an efficiency question for you. I wrote some code to analyze a report that holds over 70k records and over 400+ unique organizations to allow my supervisor to enter in year/month/date they are interested in and have it pop out the information.
The beginning of my code is:
import pandas as pd
import numpy as np
import datetime
main_data = pd.read_excel("UpdatedData.xlsx", encoding= 'utf8')
#column names from DF
epi_expose = "EpitheliumExposureSeverity"
sloughing = "EpitheliumSloughingPercentageSurface"
organization = "OrgName"
region = "Region"
date = "DeathOn"
#list storage of definitions
sl_list = ["",'None','Mild','Mild to Moderate']
epi_list= ['Moderate','Moderate to Severe','Severe']
#Create DF with four columns
df = main_data[[region, organization, epi_expose, sloughing, date]]
#filter it down to months
starting_date = datetime.date(2017,2,1)
ending_date = datetime.date(2017,2,28)
df = df[(df[date] > starting_date) & (df[date] < ending_date)]
I am then performing conditional filtering below to get counts by region and organization. It works, but is slow. Is there a more efficient way to query my DF and set up a DF that ONLY has the dates that it is supposed to sit between? Or is this the most efficient way without altering how the Database I am using is set up?
I can provide more of my code but if I filter it out by month before exporting to excel, the code runs in a matter of seconds so I am not concerned about the speed of it besides getting the correct date fields.
Thank you!

Filling data using .fillNA(), data pulled from Quandl

I've pulled some stock data from Quandl for both Crude Oil prices (WTI) and Caterpillar (CAT) price. When I concatenate the two dataframes together I'm left with some NaNs. My ultimate goal is to run a .Pearsonr() to assess the correlation (along with p-values), however I can't get Pearsonr() to work because of all the Nan's. So I'm trying to clean them up. When I use the .fillNA() function it doesn't seem to be working. I've even tried .interpolate() as well as .dropna(). None of them appear to work. Here is my working code.
import Quandl
import pandas as pd
import numpy as np
#WTI Data#
WTI_daily = Quandl.get("DOE/RWTC", collapse="daily",trim_start="1986-10-10", trim_end="1986-10-15")
WTI_daily.columns = ['WTI']
#CAT Data
CAT_daily = Quandl.get("YAHOO/CAT.6", collapse = "daily",trim_start="1986-10-10", trim_end="1986-10-15")
CAT_daily.columns = ['CAT']
#Combine Data Frames
daily_price_df = pd.concat([CAT_daily, WTI_daily], axis=1)
print daily_price_df
#Verify they are dataFrames:
def really_a_df(var):
if isinstance(var, pd.DataFrame):
print "DATAFRAME SUCCESS"
else:
print "Wahh Wahh"
return 'done'
print really_a_df(daily_price_df)
#Fill NAs
#CAN'T GET THIS TO WORK!!
daily_price_df.fillna(method='pad', limit=8)
print daily_price_df
# Try to interpolate
#CAN'T GET THIS TO WORK!!
daily_price_df.interpolate()
print daily_price_df
#Drop NAs
#CAN'T GET THIS TO WORK!!
daily_price_df.dropna(axis=1)
print daily_price_df
For what it's worth I've managed to get the function working when I create a dataframe from scratch using this code:
import pandas as pd
import numpy as np
d = {'a' : 0., 'b' : 1., 'c' : 2.,'d':None,'e':6}
d_series = pd.Series(d, index=['a', 'b', 'c', 'd','e'])
d_df = pd.DataFrame(d_series)
d_df = d_df.fillna(method='pad')
print d_df
Initially I was thinking that perhaps my data wasn't in dataframe form, but I used a simple test to confirm they are in fact dataframe. The only conclusion I that remains (in my opinion) is that it is something about the structure of the Quandl dataframe, or possibly the TimeSeries nature. Please know I'm somewhat new to python so structure answers for a begginner/novice. Any help is much appreciated!
pot shot - have you just forgotten to assign or use the inplace flag.
daily_price_df = daily_price_df.fillna(method='pad', limit=8)
OR
daily_price_df.fillna(method='pad', limit=8, inplace=True)

How to create a pivot table on extremely large dataframes in Pandas

I need to create a pivot table of 2000 columns by around 30-50 million rows from a dataset of around 60 million rows. I've tried pivoting in chunks of 100,000 rows, and that works, but when I try to recombine the DataFrames by doing a .append() followed by .groupby('someKey').sum(), all my memory is taken up and python eventually crashes.
How can I do a pivot on data this large with a limited ammount of RAM?
EDIT: adding sample code
The following code includes various test outputs along the way, but the last print is what we're really interested in. Note that if we change segMax to 3, instead of 4, the code will produce a false positive for correct output. The main issue is that if a shipmentid entry is not in each and every chunk that sum(wawa) looks at, it doesn't show up in the output.
import pandas as pd
import numpy as np
import random
from pandas.io.pytables import *
import os
pd.set_option('io.hdf.default_format','table')
# create a small dataframe to simulate the real data.
def loadFrame():
frame = pd.DataFrame()
frame['shipmentid']=[1,2,3,1,2,3,1,2,3] #evenly distributing shipmentid values for testing purposes
frame['qty']= np.random.randint(1,5,9) #random quantity is ok for this test
frame['catid'] = np.random.randint(1,5,9) #random category is ok for this test
return frame
def pivotSegment(segmentNumber,passedFrame):
segmentSize = 3 #take 3 rows at a time
frame = passedFrame[(segmentNumber*segmentSize):(segmentNumber*segmentSize + segmentSize)] #slice the input DF
# ensure that all chunks are identically formatted after the pivot by appending a dummy DF with all possible category values
span = pd.DataFrame()
span['catid'] = range(1,5+1)
span['shipmentid']=1
span['qty']=0
frame = frame.append(span)
return frame.pivot_table(['qty'],index=['shipmentid'],columns='catid', \
aggfunc='sum',fill_value=0).reset_index()
def createStore():
store = pd.HDFStore('testdata.h5')
return store
segMin = 0
segMax = 4
store = createStore()
frame = loadFrame()
print('Printing Frame')
print(frame)
print(frame.info())
for i in range(segMin,segMax):
segment = pivotSegment(i,frame)
store.append('data',frame[(i*3):(i*3 + 3)])
store.append('pivotedData',segment)
print('\nPrinting Store')
print(store)
print('\nPrinting Store: data')
print(store['data'])
print('\nPrinting Store: pivotedData')
print(store['pivotedData'])
print('**************')
print(store['pivotedData'].set_index('shipmentid').groupby('shipmentid',level=0).sum())
print('**************')
print('$$$')
for df in store.select('pivotedData',chunksize=3):
print(df.set_index('shipmentid').groupby('shipmentid',level=0).sum())
print('$$$')
store['pivotedAndSummed'] = sum((df.set_index('shipmentid').groupby('shipmentid',level=0).sum() for df in store.select('pivotedData',chunksize=3)))
print('\nPrinting Store: pivotedAndSummed')
print(store['pivotedAndSummed'])
store.close()
os.remove('testdata.h5')
print('closed')
You could do the appending with HDF5/pytables. This keeps it out of RAM.
Use the table format:
store = pd.HDFStore('store.h5')
for ...:
...
chunk # the chunk of the DataFrame (which you want to append)
store.append('df', chunk)
Now you can read it in as a DataFrame in one go (assuming this DataFrame can fit in memory!):
df = store['df']
You can also query, to get only subsections of the DataFrame.
Aside: You should also buy more RAM, it's cheap.
Edit: you can groupby/sum from the store iteratively since this "map-reduces" over the chunks:
# note: this doesn't work, see below
sum(df.groupby().sum() for df in store.select('df', chunksize=50000))
# equivalent to (but doesn't read in the entire frame)
store['df'].groupby().sum()
Edit2: Using sum as above doesn't actually work in pandas 0.16 (I thought it did in 0.15.2), instead you can use reduce with add:
reduce(lambda x, y: x.add(y, fill_value=0),
(df.groupby().sum() for df in store.select('df', chunksize=50000)))
In python 3 you must import reduce from functools.
Perhaps it's more pythonic/readable to write this as:
chunks = (df.groupby().sum() for df in store.select('df', chunksize=50000))
res = next(chunks) # will raise if there are no chunks!
for c in chunks:
res = res.add(c, fill_value=0)
If performance is poor / if there are a large number of new groups then it may be preferable to start the res as zero of the correct size (by getting the unique group keys e.g. by looping through the chunks), and then add in place.

Categories

Resources