I know that in PythonCaller processing is feature wise and each attribute can be access by feature.getAttribute().
But for my task I need all values of attribute (e.q. to check min, max, uniqueness etc).
I know that I can check those e.q. in AttributeValidator but for current task I need them in PythonCaller.
Is there some way to accomplish this using fmeobjects?
Something like getAllAttributeNames() but for let say column not line?
Thanks,
Submi
How about using a StatisticsCalculator rather than the PythonCaller?
How about using a StatisticsCalculator rather than the PythonCaller?
Â
And for T1 I need Min and Max (1 and 44) etc.Â
I know that it's easy outside of PythonCaller but since I doing other stuff there I would prefer to do those thing in PythonCaller as well. It's more question - can I work in PythonCaller let say column wise?Yes, you can use regular Python data structures such as dictionaries, lists and sets to accumulate values between features. Here's an example that returns duplicate values in attribute T1:
import fme
import fmeobjects
from collections import Counter
class FeatureProcessor(object):
    def __init__(self):
        self.T1_values = _]
       Â
    def input(self,feature):
        t1 = feature.getAttribute('T1')
        self.T1_values.append(t1)
       Â
    def close(self):
        f = fmeobjects.FMEFeature()
        T1_duplicates = k for k,v in Counter(self.T1_values).items() if v>1]
        f.setAttribute('T1_duplicates', T1_duplicates)
        self.pyoutput(f)
The PythonCaller will output one single feature containing a list attribute T1_duplicates{} that contain all the duplicates found in the T1 attribute of all the features.
Hopefully it can give you some ideas.
Yes, you can use regular Python data structures such as dictionaries, lists and sets to accumulate values between features. Here's an example that returns duplicate values in attribute T1:
import fme
import fmeobjects
from collections import Counter
class FeatureProcessor(object):
    def __init__(self):
        self.T1_values = _]
       Â
    def input(self,feature):
        t1 = feature.getAttribute('T1')
        self.T1_values.append(t1)
       Â
    def close(self):
        f = fmeobjects.FMEFeature()
        T1_duplicates = k for k,v in Counter(self.T1_values).items() if v>1]
        f.setAttribute('T1_duplicates', T1_duplicates)
        self.pyoutput(f)
The PythonCaller will output one single feature containing a list attribute T1_duplicates{} that contain all the duplicates found in the T1 attribute of all the features.
Hopefully it can give you some ideas.
Â
Thanks, I will start with Python structures. I was just wondering if there is something in fmeobjects for such things.Â
Â
Thanks, I will start with Python structures. I was just wondering if there is something in fmeobjects for such things.Â
Â
Just in case, here's the complete fmeobjects documentation where you can find all the classes and methods available to you:Â
http://docs.safe.com/fme/html/FME_Objects_Python_API/index.htmlÂ
Â
Just in case, here's the complete fmeobjects documentation where you can find all the classes and methods available to you:Â
http://docs.safe.com/fme/html/FME_Objects_Python_API/index.htmlÂ
Â
Yes, you can use regular Python data structures such as dictionaries, lists and sets to accumulate values between features. Here's an example that returns duplicate values in attribute T1:
import fme
import fmeobjects
from collections import Counter
class FeatureProcessor(object):
    def __init__(self):
        self.T1_values = _]
       Â
    def input(self,feature):
        t1 = feature.getAttribute('T1')
        self.T1_values.append(t1)
       Â
    def close(self):
        f = fmeobjects.FMEFeature()
        T1_duplicates = k for k,v in Counter(self.T1_values).items() if v>1]
        f.setAttribute('T1_duplicates', T1_duplicates)
        self.pyoutput(f)
The PythonCaller will output one single feature containing a list attribute T1_duplicates{} that contain all the duplicates found in the T1 attribute of all the features.
Hopefully it can give you some ideas.
Â
Hi submi,
If you just want to transfer all attributes from input feature to a newly created, it's as simple as:
for k in feature.getAllAttributeNames():
    newfeature.setAttribute(k,feature.getAttribute(k))
The trick is, as mentioned by others, to use the "attribute names" collection/list.
Cheers
Yes, you can use regular Python data structures such as dictionaries, lists and sets to accumulate values between features. Here's an example that returns duplicate values in attribute T1:
import fme
import fmeobjects
from collections import Counter
class FeatureProcessor(object):
    def __init__(self):
        self.T1_values = _]
       Â
    def input(self,feature):
        t1 = feature.getAttribute('T1')
        self.T1_values.append(t1)
       Â
    def close(self):
        f = fmeobjects.FMEFeature()
        T1_duplicates = k for k,v in Counter(self.T1_values).items() if v>1]
        f.setAttribute('T1_duplicates', T1_duplicates)
        self.pyoutput(f)
The PythonCaller will output one single feature containing a list attribute T1_duplicates{} that contain all the duplicates found in the T1 attribute of all the features.
Hopefully it can give you some ideas.
Thanks a lot @david_r, your example above really helped me get started with my first Python script and understand how you can store data as features pass through for use when needed etc.