Efficient operation over grouped dataframe Pandas - python

I have a very big Pandas dataframe where I need an ordering within groups based on another column. I know how to iterate over groups, do an operation on the group and union all those groups back into one dataframe however this is slow and I feel like there is a better way achieve this. Here is the input and what I want out of it. Input:
ID price
1 100.00
1 80.00
1 90.00
2 40.00
2 40.00
2 50.00
Output:
ID price order
1 100.00 3
1 80.00 1
1 90.00 2
2 40.00 1
2 40.00 2 (could be 1, doesn't matter too much)
2 50.00 3
Since this is over about 5kk records with around 250,000 IDs efficiency is important.

If speed is what you want, then the following should be pretty good, although it is a bit more complicated as it makes use of complex number sorting in numpy. This is similar to the approach used (my me) when writing the aggregate-sort method in the package numpy-groupies.
# get global sort order, for sorting by ID then price
full_idx = np.argsort(df['ID'] + 1j*df['price'])
# get min of full_idx for each ID (note that there are multiple ways of doing this)
n_for_id = np.bincount(df['ID'])
first_of_idx = np.cumsum(n_for_id)-n_for_id
# subtract first_of_idx from full_idx
rank = np.empty(len(df),dtype=int)
rank[full_idx] = arange(len(df)) - first_of_idx[df['ID'][full_idx]]
df['rank'] = rank+1
It takes 2s for 5m rows on my machine, which is about 100x faster than using groupby.rank from pandas (although I didn't actually run the pandas version with 5m rows because it would take too long; I'm not sure how #ayhan managed to do it in only 30s, perhaps a difference in pandas versions?).
If you do use this, then I recommend testing it thoroughly, as I have not.

You can use rank:
df["order"] = df.groupby("ID")["price"].rank(method="first")
df
Out[47]:
ID price order
0 1 100.0 3.0
1 1 80.0 1.0
2 1 90.0 2.0
3 2 40.0 1.0
4 2 40.0 2.0
5 2 50.0 3.0
It takes about 30s on a dataset of 5m rows with 250000 ID's (i5-3330) :
df = pd.DataFrame({"price": np.random.rand(5000000), "ID": np.random.choice(np.arange(250000), size = 5000000)})
%time df["order"] = df.groupby("ID")["price"].rank(method="first")
Wall time: 36.3 s

Related

Adding almost duplicates together and removing old duplicate in Pandas

Apologies for any inaccuracies in my wording, I'm fairly new to Python and brand new Pandas.
So I currently have a dataframe which contains about 1,000 accounts and their corresponding balance. However, some accounts exist twice in the data, once with their normal account number and once with "AM1" at the end of it. How can I create a new/edit the existing dataframe(either way works), so that 900002 and 900002AM1's balances are combined into 900002's balance and then 90002AM1 is removed from the df? Thank you. I know groupby would work (it's how I got to the current DF), but obviously I would need to be able to remove 'AM1' from all account numbers and then do :
df.groupby(['account#']).agg({'balance':'sum'}).resetindex)
Current DF:
account#
balance
900001
35.00
900002
25.00
900002AM1
25.00
900003
40.00
900004
20.00
900004AM1
10.00
Desired DF:
account#
balance
900001
35.00
900002
50.00
900003
40.00
900004
30.00
Extract number from account# column:
>>> df.groupby(df['account#'].str.extract(r'(^\d+)', expand=False)) \
.sum().reset_index()
account# balance
0 900001 35.0
1 900002 50.0
2 900003 40.0
3 900004 30.0
What str.extract does?
>>> df['account#'].str.extract(r'(^\d+)', expand=False)
0 900001
1 900002
2 900002 # <- 900002AM1
3 900003
4 900004
5 900004 # <- 900004AM1
Name: account#, dtype: object

Inserting missing numbers in dataframe

I have a program that ideally measures the temperature every second. However, in reality this does not happen. Sometimes, it skips a second or it breaks down for 400 seconds and then decides to start recording again. This leaves gaps in my 2-by-n dataframe, where ideally n = 86400 (the amount of seconds in a day). I want to apply some sort of moving/rolling average to it to get a nicer plot, but if I do that to the "raw" datafiles, the amount of data points becomes less. This is shown here, watch the x-axis. I know the "nice data" doesn't look nice yet; I'm just playing with some values.
So, I want to implement a data cleaning method, which adds data to the dataframe. I thought about it, but don't know how to implement it. I thought of it as follows:
If the index is not equal to the time, then we need to add a number, at time = index. If this gap is only 1 value, then the average of the previous number and the next number will do for me. But if it is bigger, say 100 seconds are missing, then a linear function needs to be made, which will increase or decrease the value steadily.
So I guess a training set could be like this:
index time temp
0 0 20.10
1 1 20.20
2 2 20.20
3 4 20.10
4 100 22.30
Here, I would like to get a value for index 3, time 3 and the values missing between time = 4 and time = 100. I'm sorry about my formatting skills, I hope it is clear.
How would I go about programming this?
Use merge with complete time column and then interpolate:
# Create your table
time = np.array([e for e in np.arange(20) if np.random.uniform() > 0.6])
temp = np.random.uniform(20, 25, size=len(time))
temps = pd.DataFrame([time, temp]).T
temps.columns = ['time', 'temperature']
>>> temps
time temperature
0 4.0 21.662352
1 10.0 20.904659
2 15.0 20.345858
3 18.0 24.787389
4 19.0 20.719487
The above is a random table generated with missing time data.
# modify it
filled = pd.Series(np.arange(temps.iloc[0,0], temps.iloc[-1, 0]+1))
filled = filled.to_frame()
filled.columns = ['time'] # Create a fully filled time column
merged = pd.merge(filled, temps, on='time', how='left') # merge it with original, time without temperature will be null
merged.temperature = merged.temperature.interpolate() # fill nulls linearly.
# Alternatively, use reindex, this does the same thing.
final = temps.set_index('time').reindex(np.arange(temps.time.min(),temps.time.max()+1)).reset_index()
final.temperature = final.temperature.interpolate()
>>> merged # or final
time temperature
0 4.0 21.662352
1 5.0 21.536070
2 6.0 21.409788
3 7.0 21.283505
4 8.0 21.157223
5 9.0 21.030941
6 10.0 20.904659
7 11.0 20.792898
8 12.0 20.681138
9 13.0 20.569378
10 14.0 20.457618
11 15.0 20.345858
12 16.0 21.826368
13 17.0 23.306879
14 18.0 24.787389
15 19.0 20.719487
First you can set the second values to actual time values as such:
df.index = pd.to_datetime(df['time'], unit='s')
After which you can use pandas' built-in time series operations to resample and fill in the missing values:
df = df.resample('s').interpolate('time')
Optionally, if you still want to do some smoothing you can use the following operation for that:
df.rolling(5, center=True, win_type='hann').mean()
Which will smooth with a 5 element wide Hanning window. Note: any window-based smoothing will cost you value points at the edges.
Now your dataframe will have datetimes (including date) as index. This is required for the resample method. If you want to lose the date, you can simply use:
df.index = df.index.time

Pandas select value in one dataframe, based on value in other df

df1:
GAME PLAY BET
0 (SWE, FIN) DRAW 10
1 (DEN, GER) WIN 20
2 (RUS, CZE) LOSS 30
df2:
GAME WIN DRAW LOSS
0 (SWE, FIN) 1.50 2.0 3.25
1 (DEN, GER) 2.00 2.5 2.10
2 (RUS, CZE) 1.05 2.1 10.00
I'd like to create a column "PAYOFF" in df1, for each game. Payoff is computed by fetching the actual odds (WIN/DRAW/LOSS) from df2, by multiplying that value with the "BET" from df1. For instance, for row 1, (SWE,FIN), the PLAY was a DRAW, and I need to use that value to fetch from the DRAW col in df2.
I can manage that by joining the 2 df's, and then some ugly del/rename of columns in a number of steps, but surely I'm missing some more elegant way to do that ? TIA, --Tommy
I think need lookup
df1['New']=df2.set_index('GAME').lookup(df1.GAME,df1.PLAY)
df1
Out[26]:
GAME PLAY BET New
0 (SWE,FIN) DRAW 10 2.0
1 (DEN,GER) WIN 20 2.0
2 (RUS,CZE) LOSS 30 10.0
I like Wen's solution better, but you can use
merged = pd.merge(
pd.concat([df1, pd.get_dummies(df1.PLAY)], axis=1),
df2,
on='GAME')
>>> merged.BET * (merged.DRAW_x * merged.DRAW_y + merged.WIN_x * merged.WIN_y + merged.LOSS_x * merged.LOSS_y)
0 20.0
1 40.0
2 300.0
dtype: float64

Python Pandas Running Totals with Resets

I would like to perform the following task. Given a 2 columns (good and bad) I would like to replace any rows for the two columns with a running total. Here is an example of the current dataframe along with the desired data frame.
EDIT: I should have added what my intentions are. I am trying to create equally binned (in this case 20) variable using a continuous variable as the input. I know the pandas cut and qcut functions are available, however the returned results will have zeros for the good/bad rate (needed to compute the weight of evidence and information value). Zeros in either the numerator or denominator will not allow the mathematical calculations to work.
d={'AAA':range(0,20),
'good':[3,3,13,20,28,32,59,72,64,52,38,24,17,19,12,5,7,6,2,0],
'bad':[0,0,1,1,1,0,6,8,10,6,6,10,5,8,2,2,1,3,1,1]}
df=pd.DataFrame(data=d)
print(df)
Here is an explanation of what I need to do to the above dataframe.
Roughly speaking, anytime I encounter a zero for either column, I need to use a running total for the column which is not zero to the next row which has a non-zero value for the column that contained zeros.
Here is the desired output:
dd={'AAA':range(0,16),
'good':[19,20,60,59,72,64,52,38,24,17,19,12,5,7,6,2],
'bad':[1,1,1,6,8,10,6,6,10,5,8,2,2,1,3,2]}
desired_df=pd.DataFrame(data=dd)
print(desired_df)
The basic idea of my solution is to create a column from a cumsum over non-zero values in order to get the zero values with the next non zero value into one group. Then you can use groupby + sum to get your the desired values.
two_good = df.groupby((df['bad']!=0).cumsum().shift(1).fillna(0))['good'].sum()
two_bad = df.groupby((df['good']!=0).cumsum().shift(1).fillna(0))['bad'].sum()
two_good = two_good.loc[two_good!=0].reset_index(drop=True)
two_bad = two_bad.loc[two_bad!=0].reset_index(drop=True)
new_df = pd.concat([two_bad, two_good], axis=1).dropna()
print(new_df)
bad good
0 1 19.0
1 1 20.0
2 1 28.0
3 6 91.0
4 8 72.0
5 10 64.0
6 6 52.0
7 6 38.0
8 10 24.0
9 5 17.0
10 8 19.0
11 2 12.0
12 2 5.0
13 1 7.0
14 3 6.0
15 1 2.0
This code treats your etch case of trailing zeros different from your desired output, it simple cuts it off. You'd have to add some extra code to catch that one with a different logic.
P.Tillmann. I appreciate your assistance with this. For the more advanced readers I would assume you to find this code appalling, as I do. I would be more than happy to take any recommendation which makes this more streamlined.
d={'AAA':range(0,20),
'good':[3,3,13,20,28,32,59,72,64,52,38,24,17,19,12,5,7,6,2,0],
'bad':[0,0,1,1,1,0,6,8,10,6,6,10,5,8,2,2,1,3,1,1]}
df=pd.DataFrame(data=d)
print(df)
row_good=0
row_bad=0
row_bad_zero_count=0
row_good_zero_count=0
row_out='NO'
crappy_fix=pd.DataFrame()
for index,row in df.iterrows():
if row['good']==0 or row['bad']==0:
row_bad += row['bad']
row_good += row['good']
row_bad_zero_count += 1
row_good_zero_count += 1
output_ind='1'
row_out='NO'
elif index+1 < len(df) and (df.loc[index+1,'good']==0 or df.loc[index+1,'bad']==0):
row_bad=row['bad']
row_good=row['good']
output_ind='2'
row_out='NO'
elif (row_bad_zero_count > 1 or row_good_zero_count > 1) and row['good']!=0 and row['bad']!=0:
row_bad += row['bad']
row_good += row['good']
row_bad_zero_count=0
row_good_zero_count=0
row_out='YES'
output_ind='3'
else:
row_bad=row['bad']
row_good=row['good']
row_bad_zero_count=0
row_good_zero_count=0
row_out='YES'
output_ind='4'
if ((row['good']==0 or row['bad']==0)
and (index > 0 and (df.loc[index-1,'good']!=0 or df.loc[index-1,'bad']!=0))
and row_good != 0 and row_bad != 0):
row_out='YES'
if row_out=='YES':
temp_dict={'AAA':row['AAA'],
'good':row_good,
'bad':row_bad}
crappy_fix=crappy_fix.append([temp_dict],ignore_index=True)
print(str(row['AAA']),'-',
str(row['good']),'-',
str(row['bad']),'-',
str(row_good),'-',
str(row_bad),'-',
str(row_good_zero_count),'-',
str(row_bad_zero_count),'-',
row_out,'-',
output_ind)
print(crappy_fix)

Data calculation in pandas python

I have:
A1 A2 Random data Random data2 Average Stddev
0 0.1 2.0 300 3000 1.05 1.343503
1 0.5 4.5 4500 450 2.50 2.828427
2 3.0 1.2 800 80 2.10 1.272792
3 9.0 9.0 900 90 9.00 0.000000
And would like to add a column 'ColumnX' that needs to have the values calculated as :
ColumnX = min(df['Random data']-df['Average'],df[Random data2]-
df[Stddev])/3.0*df['A2'])
I get the error:
ValueError: The truth value of a Series is ambiguous.
Your error has to do with pandas preferring bitwise operators and using the built in min function isn't going to work row wise.
A potential solution would be to make two new calculated columns then using the pandas dataframe .min method.
df['calc_col_1'] = df['Random data']-df['Average']
df['calc_col_2'] = (df['Random data2']-df['Stddev'])/(3.0*df['A2'])
df['min_col'] = df[['calc_col_1','calc_col_2']].min(axis=1)
The method min(axis=1) will find the min between the two columns by row then assigned to the new column. This way is efficient because you're using numpy vectorization, and it is easier to read.

Categories