All this is asking me to do is write a code that shows if there are any missing values where it is not the customers first order. I have provided the DataFrame. Should I use column 'Order_number" instead? Is my code wrong?
I named the DataFrame df_orders.
I thought my code would find the columns that have missing values and a greater order number than 1.
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 478967 entries, 0 to 478966
Data columns (total 6 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 order_id 478967 non-null int64
1 user_id 478967 non-null int64
2 order_number 478967 non-null int64
3 order_dow 478967 non-null int64
4 order_hour_of_day 478967 non-null int64
5 days_since_prior_order 450148 non-null float64
dtypes: float64(1), int64(5)
memory usage: 21.9 MB
None
# Are there any missing values where it's not a customer's first order?
m_v_fo= df_orders[df_orders['days_since_prior_order'].isna() > 1]
print(m_v_fo.head())
Empty DataFrame
Columns: [order_id, user_id, order_number, order_dow, order_hour_of_day,
days_since_prior_order]
Index: []
When you say .isna() you are returning a series of True or False. So that will never be > 1
Instead, try this:
m_v_fo= df_orders[df_orders['days_since_prior_order'].isna().sum() > 1]
If that doesn't solve the problem, then I'm not sure - try editing your question to add more detail and I can try again. :)
Update: I read your question again, and I think you're doing this out of order. First you need to filter on days_since_prior_order and then look for na.
m_v_fo = df_orders[df_orders['days_since_prior_order'] > 1].isna()
Related
I have two dataframes. One is excel file and another will be created by user inputs. Based on the user inputs and conditions on columns in the 1st dataframe, new columns should be added to 1st dataframe with calculations. I have wrote the code, which was successful for the test data, but the results are not coming to dataframe. Any help?
1st Dataframe:
Data columns (total 9 columns):
Column Non-Null Count Dtype
0 DDO Code 8621 non-null object
1 ULB Name 8621 non-null object
2 Dist. 8621 non-null object
3 Div. 8621 non-null object
4 Kgid No 8621 non-null int64
5 Name Of The Official 8621 non-null object
6 PRAN Number 8621 non-null float64
7 Join Date 8621 non-null datetime64[ns]
8 Present Basic 8621 non-null int64
dtypes: datetime64ns, float64(1), int64(2), object(5)
2nd Dataframe will be created by user inputs:
enter image description here
from the above data, I need to append 'n' columns based on the user inputs with loops and condition.
here is the code:
for a,b in zip(month_data.month_list, month_data.month_range):
for i,x in zip(contr_calc_new["Join Date"],contr_calc_new['Present Basic']):
if i.date().strftime('%Y-%m') == b.date().strftime('%Y-%m'):
contr_calc_new[a] = 0
else:
contr_calc_new[a] = int(((x + (x*rate)//100)*14//100))
this code is working for test data, but the results are not appending to the 1st dataframe by the calculation based on 2nd dataframe.
i need the result should be like below:
if [join date] column is equal to year & month entered by user, it must return zero, else it should return some calculation. Advance thanks for the help.
Finally I found the proper code. Thank you for your replies.
for a,b in zip(month_data.month_list, month_data.month_range):
contr_calc_new[a] = np.where(contr_calc_new['Join Date'].dt.strftime('%Y-%m') == b.date().strftime('%Y-%m'),0,((contr_calc_new['Present Basic'] + (contr_calc_new['Present Basic']*da_rate)//100)*14//100).astype(int))
I am trying to create a new pandas dataframe displayDF with 4 columns from the dataframe finalDF.
displayDF = finalDF[['False','True','RULE ID','RULE NAME']]
This command is failing with the error:
KeyError: "['False', 'True'] not in index"
However, I can see the columns "False" and "True" when I run finalDF.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 12 entries, 0 to 11
Data columns (total 6 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 rule_rec_id 12 non-null object
1 False 12 non-null int64
2 True 12 non-null int64
3 RULE ID 12 non-null object
4 RULE NAME 12 non-null object
5 RULE DESCRIPTION 12 non-null object
dtypes: int64(2), object(4)
memory usage: 672.0+ bytes
Additional Background:
I created finalDF by merging two dataframes (pivot_stackedPandasDF and dfPandaDescriptions)
finalDF = pd.merge(pivot_stackedPandasDF, dfPandaDescriptions, how='left', left_on=['rule_rec_id'], right_on=['RULE ID'])
I created pivot_stackedPandasDF with this command.
pivot_stackedPandasDF = stackedPandasDF.pivot_table(index="rule_rec_id", columns="alert_value", values="count").reset_index()
I think the root cause may be in the way I ran the .pivot_table() command.
I have a dataframe that has 3 columns and looks like this:
name date result
Anya 2021-02-13 0
Frank 2021-02-14 1
The other dataframe looks like this:
name date
Anya 2021-02-13
Frank 2021-02-14
I need to match the data types of one df to another. Because I have one additional column in df_1 I got an error. My code looks like this:
df_1.info()
Data columns (total 3 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 name 717 non-null object
1 date 717 non-null object
2 result 717 non-null int64
df_2.info()
Data columns (total 3 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 name 717 non-null object
1 date 717 non-null datetime64[ns]
# Match the primary df to secondary df
for x in df_1.columns:
df_2[x] = df_2[x].astype(df_1[x].dtypes.name)
I got an error: KeyError: 'profitable' What would be a workaround here? I need the dtypes of df_2 to be exactly the same as df_1. Thanks!
df1->that has 3 columns
df2->other dataframe
Firstly make use of boolean mask to find out those columns which are common in both dataframes:
mask=df1.columns.isin(df2.columns)
df=df1[df1.columns[mask]]
Now finally make use of astype() method:
df2=df2.astype(df.dtypes)
Or you can do all this in 1 line by:
df2=df2.astype(df1[df1.columns[df1.columns.isin(df2.columns)]].dtypes)
This question already has answers here:
Replacing blank values (white space) with NaN in pandas
(13 answers)
Closed 2 years ago.
so, I've been working with pandas in python and I got extracted data from external system with lots of spaces at the end of each column. I got an idea to use on each Series a str.strip() method with a code:
Data["DESCRIPTION"] = Data["DESCRIPTION"].str.strip()
It basically did its job but I noticed that when I check properties of data frame using I run into an issue that if in one value there were only spaces without any text then it is empty but it does not convert that scalar as null:
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18028 entries, 0 to 18027
Data columns (total 11 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 VIN 18028 non-null object
1 DESCRIPTION 18028 non-null object
2 DESCRIPTION 2 18028 non-null object
3 ENGINE 18023 non-null object
4 TRANSMISSION 18028 non-null object
5 PAINT 18028 non-null object
6 EXT_COLOR_CODE 18028 non-null object
7 EXT_COLOR_DESC 18028 non-null object
8 INT_COLOR_DESC 18028 non-null object
9 COUNTRY 18028 non-null object
10 PROD_DATE 18028 non-null object
dtypes: object(11)
memory usage: 1.5+ MB
However checking a condition if the string is empty:
Data['DESCRIPTION 2'] == ""
0 True
1 True
2 True
3 True
4 True
...
18023 True
18024 True
18025 True
18026 True
18027 True
Name: DESCRIPTION 2, Length: 18028, dtype: bool
How could I possibly convert all those as null so I could drop them using dropna() function?
I'd be grateful for any suggestions.
To remove trailing spaces and replace an empty string or records with only spaces as Nan run the below command.
Data["DESCRIPTION"].str.strip().replace(r'^\s*$', np.nan, regex=True)
Please refer to this page Replacing blank values (white space) with NaN in pandas
I am trying to get the count of unique values in a column using groupby. However I get different results if I just look at groupby results (which are printed for every group) and if I use get_group(). But the results I get is not same for the first method. What is the problem here?
print "Groupby:",bigDF[bigDF.Class == "apple"].groupby('sizeBin').customerId.nunique()
print "Selection:",bigDF[(bigDF.Class == "apple")&(bigDF.sizeBin == 0)].customerId.nunique()
print "Get group:",bigDF[bigDF.Class == "apple"].groupby('sizeBin').get_group(0).customerId.nunique()
Groupby: sizeBin
0 6
1 14
5 26
10 34
20 32
50 3
100 3
200 7
500 0
Name: customerId, dtype: int64
Selection: 34
Get group: 34
I should also note the data types, pd.info() gives me the following, so the sizeBin is a category:
<class 'pandas.core.frame.DataFrame'>
Int64Index: 224903 entries, 0 to 20616
Data columns (total 3 columns):
customerId 224903 non-null int64
Class 224903 non-null object
sizeBin 224903 non-null category
dtypes: category(1), int64(1), object(1)
memory usage: 5.4+ MB