如何遍历值是元组的字典熊猫并找到第一个True和False值

Question

I have a dictionary called weeks_adopted where when I run iteritems() and print the value , I get (example of the values for 3 keys, each key is called app_id ). The weeks_adopted dict consists of key value pairs where the key is of type <type 'str'> and the value is a <class 'pandas.core.series.Series'> where dtype is bool . Here is one example of one value , where the indices are basically the week referred to (weeks 0-13 of the year in order):

Name: app_id_str, dtype: bool
0     False
1     False
2     False
3     False
4     False
5     False
6     False
7     False
8     False
9      True
10    False
11    False
12     True
13    False
Name: app_id_str, dtype: bool
0     False
1     False
2     False
3      True
4     False
5     False
6     False
7     False
8     False
9     False
10    False
11    False
12    False
13    False
Name: app_id_str, dtype: bool
0     False
1     False
2     False
3     False
4     False
5     False
6     False
7     False
8     False
9     False
10    False
11     True
12     True
13     True

What I want to do is calculate the number of rows from the first True value right through to the first False value, for each key, obviously accounting for each cases for example in the 3rd tuple you see the first True after the first False . Basically this is to do with drop out rates - when does a user first see something (True) and then give it up (False).

In the example of the tuples above, the result should be 1, 1 and 3 in terms of adoption rate.

Here is my current basic method:

for key,value in weeks_adopted.iteritems():
    start= value.index(True)
    end = value.index(False)
    adoption=end-start
    weeks_adopted[key] = adoption

However I get this error even with this method:

TypeError                                 Traceback (most recent call last)
<ipython-input-32-608c4f533e54> in <module>()
     19 for key,value in weeks_adopted.iteritems():
     20     print value
---> 21     start= value.index(True)
     22     end = value.index(False)
     23     adoption=end-start

TypeError: 'Int64Index' object is not callable

In the answer, please could you help me in what other checks I need to be doing to find the first True and first Last value? I am presuming this type of loop is a common one for many situations?

Answer 1

you can try this:

def calc_adoption(ts):
    true_index = ts[ts].index
    if len(true_index) == 0:
        return 0
    first_true_index = true_index[0]
    false_index = ts.index.difference(true_index)
    false_index = false_index[false_index > first_true_index]
    if len(false_index) == 0:
        return 14 - first_true_index
    return false_index[0] - first_true_index

adopted_weeks = {k: calc_adoption(v) for k, v in weeks_adopted.iteritems()}