Return True only when first numeric entry is found & False in all other cases-CodePudding

The dataframe I am dealing with looks like the table below:

 COLUMN-A      COLUMN-B           COLUMN-C            COLUMN-D
2005-12-23  2.78229429977895    2.59054751268432    
2005-12-28  2.77990953370726    2.59625529291923    
2005-12-29  2.77770141742004    2.60175855794512    
2005-12-30  2.77565686568447    2.60706465870293    
2006-01-03  2.78676377607689    2.61845788272621    
2006-01-04  2.79415905904631    2.62804815466004    
2006-01-05  2.79233986786484    2.63311058575101    
2006-01-06  2.79065543181717    2.63799172343874    
2006-01-09  2.7876513234596 2.64200075091549    
2006-01-10  2.78342529650764    2.64516894228885    
2006-01-11  2.77951230901599    2.64822370776439    
2006-01-12  2.77877806345801    2.65256358425937    
2006-01-13  2.78965376857357    2.66232574953289    
2006-01-16  2.81417572440332    2.67871384606613    
2006-01-17  2.83688123723998    2.69451541833616    
2006-01-18  2.84923078073203    2.70556518000894    
2006-01-19  2.854887762274  2.71343113557577    
2006-01-20  2.86012570781281    2.72101563266667    
2006-01-23  2.8620867671879 2.72693465617535    
2006-01-24  2.85668033821582    2.72915676427006    
2006-01-25  2.85311883059988    2.7319963852241 
2006-01-27  2.84982113851717    2.73473442527192    
2006-01-30  2.84098994077245    2.73458665290639    
2006-01-31  2.83281290615161    2.73444416615124    
2006-02-01  2.82235268854652    2.73291291585375    
2006-02-02  2.79821544736977    2.72446373657389    2.31735945722146
2006-02-03  2.7903180053127 2.72328924609567    2.32165937425023
2006-02-06  2.78300555917914    2.72215675685381    2.32590335299919
2006-02-07  2.77912366526979    2.72245848891773    2.33053900014161
2006-02-08  2.77552931914827    2.72274943166327    2.33511466419111

I'm trying to write logic to return True where COLUMN-D had it's first numeric entry & False in all other cases

Here's the logic I've written which is throwing the error - ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all()

Code

import pandas as pd

    def has_trail_started(df, df_key):
        return (~pd.isnull(df[df_key])) & (pd.isnull(df[df_key].shift()))
    
    if (has_trail_started(data, 'COLUMN-D') and data['has_changed_status']):
       // Logic

Please could I get some help to rectify the problem?

CodePudding user response：

Your function returns series, which can't be interpreted as bool for purposes of if statement. But you can add the "trail start info" to the df as follows:

def has_trail_started(df, df_key):
    df["has_trail_started"] = (~pd.isnull(df[df_key])) & (pd.isnull(df[df_key].shift()))

has_trail_started(data, 'COLUMN-D')

Then df looks like this:

      COLUMN-A  COLUMN-B  COLUMN-C  COLUMN-D  has_trail_started
0   2005-12-23  2.782294  2.590548       NaN              False
1   2005-12-28  2.779910  2.596255       NaN              False
2   2005-12-29  2.777701  2.601759       NaN              False
3   2005-12-30  2.775657  2.607065       NaN              False
4   2006-01-03  2.786764  2.618458       NaN              False
5   2006-01-04  2.794159  2.628048       NaN              False
6   2006-01-05  2.792340  2.633111       NaN              False
7   2006-01-06  2.790655  2.637992       NaN              False
8   2006-01-09  2.787651  2.642001       NaN              False
9   2006-01-10  2.783425  2.645169       NaN              False
10  2006-01-11  2.779512  2.648224       NaN              False
11  2006-01-12  2.778778  2.652564       NaN              False
12  2006-01-13  2.789654  2.662326       NaN              False
13  2006-01-16  2.814176  2.678714       NaN              False
14  2006-01-17  2.836881  2.694515       NaN              False
15  2006-01-18  2.849231  2.705565       NaN              False
16  2006-01-19  2.854888  2.713431       NaN              False
17  2006-01-20  2.860126  2.721016       NaN              False
18  2006-01-23  2.862087  2.726935       NaN              False
19  2006-01-24  2.856680  2.729157       NaN              False
20  2006-01-25  2.853119  2.731996       NaN              False
21  2006-01-27  2.849821  2.734734       NaN              False
22  2006-01-30  2.840990  2.734587       NaN              False
23  2006-01-31  2.832813  2.734444       NaN              False
24  2006-02-01  2.822353  2.732913       NaN              False
25  2006-02-02  2.798215  2.724464  2.317359               True
26  2006-02-03  2.790318  2.723289  2.321659              False
27  2006-02-06  2.783006  2.722157  2.325903              False
28  2006-02-07  2.779124  2.722458  2.330539              False
29  2006-02-08  2.775529  2.722749  2.335115              False

Now you can apply some logic based on this new bool like this:

data["extra_logic"] = data["has_trail_started"].apply(lambda x: "yay" if x else "boo")

Which will add a new column with values being a function of has_trail_started flag.