Develop Reference

Efficient row comparison in pandas dataframe on incomplete data

I work on an incomplete data that also has doubles and I need to clear it from doubles, choosing complete rows if available.
For example: that's how the data look
I need to search trough each row to see whether it's a double (has a 'rank'>1), and whether if it is incomplete itself, but has some complete doubles.
I'll explain now:
not every row with the 'rank' = 1 has a date in it (it is crutial),
but some of them have doubles ('rank'>1) which has a date.
not every row has a double. And if it doesn't have a date in it, that's ok.
So, I need to find the double with the date if it does exist, and rewrite it to the row with the rank 1 (or delete an incomplete first row)
In the end I need to have a DataFrame with no doubles and as much dates as available.
There's my code with EXTREMELY inefficient iterative loop, but I don't know how to rewrite it with vectorization or .apply() method:
def test_func(dataframe):
df = dataframe
df.iloc[0:0]
for i in range(0, dataframe.shape[0]):
if dataframe.iloc[i]['rank'] == 1:
temp_row = dataframe.iloc[i]
elif ((dataframe.iloc[i+1]['rank']>1)&
(pd.isna(dataframe.iloc[i]['date'])
&(~pd.isna(dataframe.iloc[i+1]['date'])))):
temp_row = dataframe.iloc[i+1]
df.loc[i] = temp_row
return df
Hope to find some help! From Russia with love xo.

Assuming that you are grouping by phone, and you are interested in populating missing dates, then you can use backwards fill and group by, which will fill the missing dates with the next available not null date within the group.
test_df['date'] = test_df.groupby(['phone'])['date'].apply(lambda x: x.bfill())
if you need to populate other missing data, just replace 'date' with the relevant column name

For Python Pandas, how to implement a "running" check of 2 rows against the previous 2 rows?

[updated with expected outcome]
I'm trying to implement a "running" check where I need the sum and mean of two rows to be more than the previous 2 rows.
Referring to the dataframe (copied into spreadsheet) below, I'm trying code out a function where if the mean of those two orange cells is more than the blue cells, the function will return true for row 8, under a new column called 'Cond11'. The dataframe here is historical, so all rows are available.
Note that that Rows column is added in the spreadsheet, easier for me to reference the rows here.
I have been using .rolling to refer to the current row + whatever number of rows to refer to, or using shift(1) to refer to the previous row.
df.loc[:, ('Cond9')] = df.n.rolling(4).mean() >= 30
df.loc[:, ('Cond10')] = df.a > df.a.shift(1)
I'm stuck here... how to I do this 2 rows vs the previous 2 rows? Please advise!
The 2nd part of this question: I have another function that checks the latest rows in the dataframe for the same condition above. This function is meant to be used in real-time, when new data is streaming into the dataframe and the function is supposed to check the latest rows only.
Can I check if the following code works to detect the same conditions above?
cond11 = candles.n[-2:-1].sum() > candles.n[-4:-3].sum()

I believe this solves your problem:
df.rolling(4).apply(lambda rows: rows[0] + rows[1] < rows[2] + rows[3])
The first 3 rows will be NaNs but you did not define what you would like to happen there.
As for the second part, to be able to produce this condition live for new data you just have to prepend the last 3 rows of your current data and then apply the same process to it:
pd.concat([df[-3:], df])

Python - How to optimize code to run faster? (lots of for loops in DataFrame)

I have a code that works with an excel file (SAP Download) quite extensively (data transformation and calculation steps).
I need to loop through all the lines (couple thousand rows) a few times. I have written a code prior that adds DataFrame columns separately, so I could do everything in one for loop that was of course quite quick, however, I had to change data source that meant change in raw data structure.
The raw data structure has 1st 3 rows empty, then a Title row comes with column names, then 2 rows empty, and the 1st column is also empty. I decided to wipe these, and assign column names and make them headers (steps below), however, since then, separately adding column names and later calculating everything in one for statement does not fill data to any of these specific columns.
How could i optimize this code?
I have deleted some calculation steps since they are quite long and make code part even less readable
#This function adds new column to the dataframe
def NewColdfConverter(*args):
for i in args:
dfConverter[i] = '' #previously used dfConverter[i] = NaN
#This function creates dataframe from excel file
def DataFrameCreator(path,sheetname):
excelFile = pd.ExcelFile(path)
global readExcel
readExcel = pd.read_excel(excelFile,sheet_name=sheetname)
#calling my function to create dataframe
DataFrameCreator(filePath,sheetName)
dfConverter = pd.DataFrame(readExcel)
#dropping NA values from Orders column (right now called Unnamed)
dfConverter.dropna(subset=['Unnamed: 1'], inplace=True)
#dropping rows and deleting other unnecessary columns
dfConverter.drop(dfConverter.head(1).index, inplace=True)
dfConverter.drop(dfConverter.columns[[0,11,12,13,17,22,23,48]], axis = 1,inplace = True)
#renaming columns from Unnamed 1: etc to proper names
dfConverter = dfConverter.rename(columns={Unnamed 1:propername1 Unnamed 2:propername2 etc.})
#calling new column function -> this Day column appears in the 1st for loop
NewColdfConverter("Day")
#example for loop that worked prior, but not working since new dataset and new header/column steps added:
for i in range(len(dfConverter)):
#Day column-> floor Entry Date -1, if time is less than 5:00:00
if(dfConverter['Time'][i] <= time(hour=5,minute=0,second=0)):
dfConverter['Day'][i] = pd.to_datetime(dfConverter['Entry Date'][i])-timedelta(days=1)
else:
dfConverter['Day'][i] = pd.to_datetime(dfConverter['Entry Date'][i])
Problem is, there are many columns that build on one another, so I cannot get them in one for loop, for instance in below example I need to calculate reqsWoSetUpValue, so I can calculate requirementsValue, so I can calculate otherReqsValue, but I'm not able to do this within 1 for loop by assigning the values to the dataframecolumn[i] row, because the value will just be missing, like nothing happened.
(dfsorted is the same as dfConverter, but a sorted version of it)
#example code of getting reqsWoSetUpValue
for i in range(len(dfSorted)):
reqsWoSetUpValue[i] = #calculationsteps...
#inserting column with value
dfSorted.insert(49,'Reqs wo SetUp',reqsWoSetUpValue)
#getting requirements value with previously calculated Reqs wo SetUp column
for i in range(len(dfSorted)):
requirementsValue[i] = #calc
dfSorted.insert(50,'Requirements',requirementsValue)
#Calculating Other Reqs value with previously calculated Requirements column.
for i in range(len(dfSorted)):
otherReqsValue[i] = #calc
dfSorted.insert(51,'Other Reqs',otherReqsValue)
Anyone have a clue, why I cannot do this in 1 for loop anymore by 1st adding all columns by the function, like:
NewColdfConverter('Reqs wo setup','Requirements','Other reqs')
#then in 1 for loop:
for i in range(len(dfsorted)):
dfSorted['Reqs wo setup'] = #calculationsteps
dfSorted['Requirements'] = #calculationsteps
dfSorted['Other reqs'] = #calculationsteps
Thank you

General comment: How to identify bottlenecks
To get started, you should try to identify which parts of the code are slow.
Method 1: time code sections using the time package
Wrap blocks of code in statements like this:
import time
t = time.time()
# do something
print("time elapsed: {:.1f} seconds".format(time.time() - t))
Method 2: use a profiler
E.g. Spyder has a built-in profiler. This allows you to check which operations are most time consuming.
Vectorize your operations
Your code will be orders of magnitude faster if you vectorize your operations. It looks like your loops are all avoidable.
For example, rather than calling pd.to_datetime on every row separately, you should call it on the entire column at once
# slow (don't do this):
for i in range(len(dfConverter)):
dfConverter['Day'][i] = pd.to_datetime(dfConverter['Entry Date'][i])
# fast (do this instead):
dfConverter['Day'] = pd.to_datetime(dfConverter['Entry Date'])
If you want to perform an operation on a subset of rows, you can also do this in a vectorized operation by using loc:
mask = dfConverter['Time'] <= time(hour=5,minute=0,second=0)
dfConverter.loc[mask,'Day'] = pd.to_datetime(dfConverter.loc[mask,'Entry Date']) - timedelta(days=1)

Not sure this would improve performance, but you could calculate the dependent columns at the same time row by row with DataFrame.iterrows()
for index, data in dfSorted.iterrows():
dfSorted['Reqs wo setup'][index] = #calculationsteps
dfSorted['Requirements'][index] = #calculationsteps
dfSorted['Other reqs'][index] = #calculationsteps

How to return the string of a header based on the max value of a cell in Openpyxl

Good morning guys! quick question for Openpyxl:
I am working with Python editing a xlsx document and generating various stats. Part of my script is to generate max values of a cell range :
temp_list=[]
temp_max=[]
for row in sheet.iter_rows(min_row=3, min_col=10, max_row=508, max_col=13):
print(row)
for cell in row:
temp_list.append(cell.value)
print(temp_list)
temp_max.append(max(temp_list))
temp_list=[]
I would also like to be able to print the string of the header of the column that contains the max value for the cell range desired. My data structure looks like this :
Any idea on how to do so?
Thanks!

This seems like a typical INDEX/MATCH Excel problem.
Have you tried retrieving the index for the max value in each temp_list?
You can use a function like numpy.argmax() to get the index of your max value within your "temp_list" array, then use this index to locate the header and append the string to a new list called, say, "max_headers" which contains all the header strings in order of appearance.
It would look something like this
for cell in row:
temp_list.append(cell.value)
i_max = np.argmax(temp_list)
max_headers.append(cell(row = 1, column = i_max).value)
And so on and so forth. Of course, for that to work, your temp_list should be a numpy array instead of a simple python list, and the max_headers list would have to be defined.

First, Thanks Bernardo for the hint. I found a decently working solution but still have a little issue. Perhaps someone can be of assistance.
Let me amend my initial statement : here is the code I am working with now :
temp_list=[]
headers_list=[]
for row in sheet.iter_rows(min_row=3, min_col=27, max_row=508, max_col=32): #Index starts at 1 // Here we set the rows/columns containing the data to be analyzed
for cell in row:
temp_list.append(cell.value)
for cell in row:
if cell.value == max(temp_list):
print(str(cell.column))
print(cell.value)
print(sheet.cell(row=1, column=cell.column).value)
headers_list.append(sheet.cell(row=1,column=cell.column).value)
else:
print('keep going.')
temp_list = []
This formula works, but has a little issue : If, for instance, a row has the same value twice (ie : 25,9,25,8,9), this loop will print 2 headers instead of one. My question is :
how can I get this loop to take in account only the first match of a max value in a row?

You probably want something like this:
headers = [c for c in next(ws.iter_rows(min_col=27, max_col=32, min_row=1, max_row=1, values_only=True))]
for row in ws.iter_rows(min_row=3, min_col=27, max_row=508, max_col=32, values_only=True):
mx = max(row)
idx = row.index(mx)
col = headers[idx]

Pandas For Loop, If String Is Present In ColumnA Then ColumnB Value = X

I'm pulling Json data from the Binance REST API, after formatting I'm left with the following...
I have a dataframe called Assets with 3 columns [Asset,Amount,Location],
['Asset'] holds ticker names for crypto assets e.g.(ETH,LTC,BNB).
However when all or part of that asset has been moved to 'Binance Earn' the strings are returned like this e.g.(LDETH,LDLTC,LDBNB).
['Amount'] can be ignored for now.
['Location'] is initially empty.
I'm trying to set the value of ['Location'] to 'Earn' if the string in ['Asset'] includes 'LD'.
This is how far I got, but I can't remember how to apply the change to only the current item, it's been ages since I've used Pandas or for loops.
And I'm only able to apply it to the entire column rather than the row iteration.
for Row in Assets['Asset']:
if Row.find('LD') == 0:
print('Earn')
Assets['Location'] = 'Earn' # <----How to apply this to current row only
else:
print('???')
Assets['Location'] = '???' # <----How to apply this to current row only
The print statements work correctly, but currently the whole column gets populated with the same value (whichever was last) as you might expect.
So (LDETH,HOT,LDBTC) returns ('Earn','Earn','Earn') rather than the desired ('Earn','???','Earn')
Any help would be appreciated...

np.where() fits here. If the Asset starts with LD, then return Earn, else return ???:
Assets['Location'] = np.where(Assets['Asset'].str.startswith('LD'), 'Earn', '???')

You could run a lambda in df.apply to check whether 'LD' is in df['Asset']:
df['Location'] = df['Asset'].apply(lambda x: 'Earn' if 'LD' in x else None)

One possible solution:
def get_loc(row):
asset = row['Asset']
if asset.find('LD') == 0:
print('Earn')
return 'Earn'
print('???')
return '???'
Assets['Location'] = Assets.apply(get_loc, axis=1)
Note, you should almost never iterate over a pandas dataframe or series.