I want to do some complex calculations in pandas while referencing previous values (basically I'm calculating row by row). However the loops take forever and I wanted to know if there was a faster way. Everybody keeps mentioning using shift but I don't understand how that would even work.
df = pd.DataFrame(index=range(500)
df["A"]= 2
df["B"]= 5
df["A"][0]= 1
for i in range(len(df):
if i != 0: df['A'][i] = (df['A'][i-1] / 3) - df['B'][i-1] + 25
numpy_ext can be used for expanding calculations
pandas-rolling-apply-using-multiple-columns for reference
I have also included a simpler calc to demonstrate behaviour in simpler way
df = pd.DataFrame(index=range(5000))
df["A"]= 2
df["B"]= 5
df["A"][0]= 1
import numpy_ext as npe
# for i in range(len(df):
# if i != 0: df['A'][i] = (df['A'][i-1] / 3) - df['B'][i-1] + 25
# SO example - function of previous values in A and B
def f(A,B):
r = np.sum(A[:-1]/3) - np.sum(B[:-1] + 25) if len(A)>1 else A[0]
return r
# much simpler example, sum of previous values
def g(A):
return np.sum(A[:-1])
df["AB_combo"] = npe.expanding_apply(f, 1, df["A"].values, df["B"].values)
df["A_running"] = npe.expanding_apply(g, 1, df["A"].values)
print(df.head(10).to_markdown())
sample output
A
B
AB_combo
A_running
0
1
5
1
0
1
2
5
-29.6667
1
2
2
5
-59
3
3
2
5
-88.3333
5
4
2
5
-117.667
7
5
2
5
-147
9
6
2
5
-176.333
11
7
2
5
-205.667
13
8
2
5
-235
15
9
2
5
-264.333
17
I am new to python and the last time I coded was in the mid-80's so I appreciate your patient help.
It seems .rolling(window) requires the window to be a fixed integer. I need a rolling window where the window or lookback period is dynamic and given by another column.
In the table below, I seek the Lookbacksum which is the rolling sum of Data as specified by the Lookback column.
d={'Data':[1,1,1,2,3,2,3,2,1,2],
'Lookback':[0,1,2,2,1,3,3,2,3,1],
'LookbackSum':[1,2,3,4,5,8,10,7,8,3]}
df=pd.DataFrame(data=d)
eg:
Data Lookback LookbackSum
0 1 0 1
1 1 1 2
2 1 2 3
3 2 2 4
4 3 1 5
5 2 3 8
6 3 3 10
7 2 2 7
8 1 3 8
9 2 1 3
You can create a custom function for use with df.apply, eg:
def lookback_window(row, values, lookback, method='sum', *args, **kwargs):
loc = values.index.get_loc(row.name)
lb = lookback.loc[row.name]
return getattr(values.iloc[loc - lb: loc + 1], method)(*args, **kwargs)
Then use it as:
df['new_col'] = df.apply(lookback_window, values=df['Data'], lookback=df['Lookback'], axis=1)
There may be some corner cases but as long as your indices align and are unique - it should fulfil what you're trying to do.
here is one with a list comprehension which stores the index and value of the column df['Lookback'] and the gets the slice by reversing the values and slicing according to the column value:
df['LookbackSum'] = [sum(df.loc[:e,'Data'][::-1].to_numpy()[:i+1])
for e,i in enumerate(df['Lookback'])]
print(df)
Data Lookback LookbackSum
0 1 0 1
1 1 1 2
2 1 2 3
3 2 2 4
4 3 1 5
5 2 3 8
6 3 3 10
7 2 2 7
8 1 3 8
9 2 1 3
An exercise in pain, if you want to try an almost fully vectorized approach. Sidenote: I don't think it's worth it here. At all.
Inspired by Divakar's answer here
Given:
import numpy as np
import pandas as pd
d={'Data':[1,1,1,2,3,2,3,2,1,2],
'Lookback':[0,1,2,2,1,3,3,2,3,1],
'LookbackSum':[1,2,3,4,5,8,10,7,8,3]}
df=pd.DataFrame(data=d)
Using the function from Divakar's answer, but slightly modified
from skimage.util.shape import view_as_windows as viewW
def strided_indexing_roll(a, r, fill_value=np.nan):
# Concatenate with sliced to cover all rolls
p = np.full((a.shape[0],a.shape[1]-1),fill_value)
a_ext = np.concatenate((p,a,p),axis=1)
# Get sliding windows; use advanced-indexing to select appropriate ones
n = a.shape[1]
return viewW(a_ext,(1,n))[np.arange(len(r)), -r + (n-1),0]
Now, we just need to prepare a 2d array for the data and independently shift the rows according to our desired lookback values.
arr = df['Data'].to_numpy().reshape(1, -1).repeat(len(df), axis=0)
shifter = np.arange(len(df) - 1, -1, -1) #+ d['Lookback'] - 1
temp = strided_indexing_roll(arr, shifter, fill_value=0)
out = strided_indexing_roll(temp, (len(df) - 1 - df['Lookback'])*-1, 0).sum(-1)
Output:
array([ 1, 2, 3, 4, 5, 8, 10, 7, 8, 3], dtype=int64)
We can then just assign it back to the dataframe as needed and check.
df['out'] = out
#output:
Data Lookback LookbackSum out
0 1 0 1 1
1 1 1 2 2
2 1 2 3 3
3 2 2 4 4
4 3 1 5 5
5 2 3 8 8
6 3 3 10 10
7 2 2 7 7
8 1 3 8 8
9 2 1 3 3
I have a dataset df of object components in 3-d space - each ID represents an object which has various components:
ID Comp x y z
A 1 2 2 1
A 2 2 1 -1
A 3 -10 1 -10
A 4 -1 3 -5
B 1 3 0 0
B 2 3 0 -5
...
I would like to loop through each ID, using a clustering technique in sklearn to create clusters of components (Comp) based on each component's (x,y,z) co-ordinates - to achieve something like this:
ID Comp x y z cluster
A 1 2 2 1 1
A 2 2 1 -1 1
A 3 -10 1 -10 2
A 4 -1 3 -5 3
B 1 3 0 0 1
B 2 3 0 -5 1
...
As an example - ID:A,Comp:1 is incluster1, whereasID:A, Comp:4 is in cluster 3. (I plan to then concatenate ID and cluster later).
I'm having no luck with the following groupby + apply:
from sklearn.cluster import AffinityPropagation
ap = AffinityPropagation()
df['cluster']=df.groupby(['ID','Comp']).apply(lambda x: ap.fit_predict(np.array([x.x,x.y,x.z]).T))
I could brute-force it by using a for loop over the ID but my dataset is large (~ 150k ID) and I'm worried about resource and time constraints. Any help would be great!
IIUC, I think you could try something like this:
def ap_fit_pred(x):
ap = AffinityPropagation()
return pd.Series(ap.fit_predict(x.loc[:,['x','y','z']]))
df['cluster'] = df.groupby('ID').apply(ap_fit_pred).reset_index(drop=True)
How to calculate amounts that row values greater than a specific value in pandas?
For example, I have a Pandas DataFrame dff. I want to count row values greater than 0.
dff = pd.DataFrame(np.random.randn(9,3),columns=['a','b','c'])
dff
a b c
0 -0.047753 -1.172751 0.428752
1 -0.763297 -0.539290 1.004502
2 -0.845018 1.780180 1.354705
3 -0.044451 0.271344 0.166762
4 -0.230092 -0.684156 -0.448916
5 -0.137938 1.403581 0.570804
6 -0.259851 0.589898 0.099670
7 0.642413 -0.762344 -0.167562
8 1.940560 -1.276856 0.361775
I am using an inefficient way. How to be more efficient?
dff['count'] = 0
for m in range(len(dff)):
og = 0
for i in dff.columns:
if dff[i][m] > 0:
og += 1
dff['count'][m] = og
dff
a b c count
0 -0.047753 -1.172751 0.428752 1
1 -0.763297 -0.539290 1.004502 1
2 -0.845018 1.780180 1.354705 2
3 -0.044451 0.271344 0.166762 2
4 -0.230092 -0.684156 -0.448916 0
5 -0.137938 1.403581 0.570804 2
6 -0.259851 0.589898 0.099670 2
7 0.642413 -0.762344 -0.167562 1
8 1.940560 -1.276856 0.361775 2
You can create a boolean mask of your DataFrame, that is True wherever a value is greater than your threshold (in this case 0), and then use sum along the first axis.
dff.gt(0).sum(1)
0 1
1 1
2 2
3 2
4 0
5 2
6 2
7 1
8 2
dtype: int64
Having a DataFrame with the following column:
df['A'] = [1,1,1,0,1,1,1,1,0,1]
What would be the best vectorized way to control the length of "1"-series by some limiting value? Let's say the limit is 2, then the resulting column 'B' must look like:
A B
0 1 1
1 1 1
2 1 0
3 0 0
4 1 1
5 1 1
6 1 0
7 1 0
8 0 0
9 1 1
One fully-vectorized solution is to use the shift-groupby-cumsum-cumcount combination1 to indicate where consecutive runs are shorter than 2 (or whatever limiting value you like). Then, & this new boolean Series with the original column:
df['B'] = ((df.groupby((df.A != df.A.shift()).cumsum()).cumcount() <= 1) & df.A)\
.astype(int) # cast the boolean Series back to integers
This produces the new column in the DataFrame:
A B
0 1 1
1 1 1
2 1 0
3 0 0
4 1 1
5 1 1
6 1 0
7 1 0
8 0 0
9 1 1
1 See the pandas cookbook; the section on grouping, "Grouping like Python’s itertools.groupby"
Another way (checking if previous two are 1):
In [443]: df = pd.DataFrame({'A': [1,1,1,0,1,1,1,1,0,1]})
In [444]: limit = 2
In [445]: df['B'] = map(lambda x: df['A'][x] if x < limit else int(not all(y == 1 for y in df['A'][x - limit:x])), range(len(df)))
In [446]: df
Out[446]:
A B
0 1 1
1 1 1
2 1 0
3 0 0
4 1 1
5 1 1
6 1 0
7 1 0
8 0 0
9 1 1
If you know that the values in the series will all be either 0 or 1, I think you can use a little trick involving convolution. Make a copy of your column (which need not be a Pandas object, it can just be a normal Numpy array)
a = df['A'].as_matrix()
and convolve it with a sequence of 1's that is one longer than the cutoff you want, then chop off the last cutoff elements. E.g. for a cutoff of 2, you would do
long_run_count = numpy.convolve(a, [1, 1, 1])[:-2]
The resulting array, in this case, gives the number of 1's that occur in the 3 elements prior to and including that element. If that number is 3, then you are in a run that has exceeded length 2. So just set those elements to zero.
a[long_run_count > 2] = 0
You can now assign the resulting array to a new column in your DataFrame.
df['B'] = a
To turn this into a more general method:
def trim_runs(array, cutoff):
a = numpy.asarray(array)
a[numpy.convolve(a, numpy.ones(cutoff + 1))[:-cutoff] > cutoff] = 0
return a