I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
df_test.pkl (1.14 KB)
···
Best,
Anubhav
I’ll take a look shortly.
Best,
Joey
···
Best,
Anubhav
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
···
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
···
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
···
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Sure, I will cite all the packages I have been using from the hackingmaterials project.
I have one comment/feature request for this featurizer. In most cases, addition and subtraction operations are limited to columns with same units for the descriptors to be physical. My naive way of dealing with this was to pass another vector with the same length as the length of the passed column names with numeric flags to restrict addition and subtraction operations to the ones with same units. I am sure there are better ways to deal with this.
Regards
···
On Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Hi,
I had question about the combo_function feature. The default is np.prod. I wanted to have addtion, subtraction and division as my combo function. Since np.prod works I thought it would be straight forward to use np.add, np.subtract and np.divide. But I get the following error for all cases.
calc=FunctionFeaturizer(expressions=[“x*x”],multi_feature_depth=2,combo_function=np.divide)
calc.exp_dict
<details class='elided'>
<summary title='Show trimmed content'>···</summary>
--------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-45-de1a409f966e> in <module>()
----> 1 calc.exp_dict
~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in exp_dict(self)
72 [(n, generate_expressions_combinations(self.expressions, n,
73 self.combo_function))
---> 74 for n in range(1, self.multi_feature_depth+1)])
75
76 def featurize_dataframe(self, df, col_id, ignore_errors=False,
~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in <listcomp>(.0)
72 [(n, generate_expressions_combinations(self.expressions, n,
73 self.combo_function))
---> 74 for n in range(1, self.multi_feature_depth+1)])
75
76 def featurize_dataframe(self, df, col_id, ignore_errors=False,
~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in generate_expressions_combinations(expressions, combo_depth, combo_function)
233 # Get all permutations of each set
234 for exp_perm in itertools.permutations(exp_set):
--> 235 combo_exps.append(combo_function(exp_perm))
236
237 # Filter for unique combinations, also remove identity
ValueError: invalid number of arguments
``
Could anyone suggest how to use the combo_function? An example would really help!
Regards,
Jonayat
On Friday, April 27, 2018 at 7:01:57 PM UTC-4, ASM Jonayat wrote:
Sure, I will cite all the packages I have been using from the hackingmaterials project.
I have one comment/feature request for this featurizer. In most cases, addition and subtraction operations are limited to columns with same units for the descriptors to be physical. My naive way of dealing with this was to pass another vector with the same length as the length of the passed column names with numeric flags to restrict addition and subtraction operations to the ones with same units. I am sure there are better ways to deal with this.
Regards
On Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
The combo_function must be a function that takes a list of inputs, which np.prod does, but np.add, np.divide, and np.subtract do not. You can use np.sum for example, or write a lambda function for the others. I’m updating the docstring to emphasize this.
···
On Saturday, April 28, 2018 at 5:42:19 PM UTC-7, ASM Jonayat wrote:
Hi,
I had question about the combo_function feature. The default is np.prod. I wanted to have addtion, subtraction and division as my combo function. Since np.prod works I thought it would be straight forward to use np.add, np.subtract and np.divide. But I get the following error for all cases.
calc=FunctionFeaturizer(expressions=[“x*x”],multi_feature_depth=2,combo_function=np.divide)
calc.exp_dict-------------------------------------------------------------------------- ValueError Traceback (most recent call last) <ipython-input-45-de1a409f966e> in <module>() ----> 1 calc.exp_dict ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in exp_dict(self) 72 [(n, generate_expressions_combinations(self.expressions, n, 73 self.combo_function)) ---> 74 for n in range(1, self.multi_feature_depth+1)]) 75 76 def featurize_dataframe(self, df, col_id, ignore_errors=False, ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in <listcomp>(.0) 72 [(n, generate_expressions_combinations(self.expressions, n, 73 self.combo_function)) ---> 74 for n in range(1, self.multi_feature_depth+1)]) 75 76 def featurize_dataframe(self, df, col_id, ignore_errors=False, ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/function.py in generate_expressions_combinations(expressions, combo_depth, combo_function) 233 # Get all permutations of each set 234 for exp_perm in itertools.permutations(exp_set): --> 235 combo_exps.append(combo_function(exp_perm)) 236 237 # Filter for unique combinations, also remove identity ValueError: invalid number of arguments
``
Could anyone suggest how to use the combo_function? An example would really help!
Regards,
JonayatOn Friday, April 27, 2018 at 7:01:57 PM UTC-4, ASM Jonayat wrote:
Sure, I will cite all the packages I have been using from the hackingmaterials project.
I have one comment/feature request for this featurizer. In most cases, addition and subtraction operations are limited to columns with same units for the descriptors to be physical. My naive way of dealing with this was to pass another vector with the same length as the length of the passed column names with numeric flags to restrict addition and subtraction operations to the ones with same units. I am sure there are better ways to deal with this.
Regards
On Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Just wanted to know if you have the reference for matminer now.
Regards,
Jonayat
···
On Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
Hi,
Thanks for checking up!
We have a paper that is accepted but not yet published in Computational Materials Science. I will update the thread again when the paper is actually published / available online.
Best,
Anubhav
···
On Sun, May 13, 2018 at 9:43 AM, ASM Jonayat [email protected] wrote:
Just wanted to know if you have the reference for matminer now.
Regards,
JonayatOn Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/68d617fd-9410-45be-8f20-3c74394e53ee%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
Hi,
The matminer paper is now published, please see:
https://authors.elsevier.com/a/1X6VX3In-uhJaK
Thanks for your support!
Anubhav
···
On Monday, May 14, 2018 at 10:18:48 AM UTC-7, Anubhav Jain wrote:
Hi,
Thanks for checking up!
We have a paper that is accepted but not yet published in Computational Materials Science. I will update the thread again when the paper is actually published / available online.
Best,
Anubhav
On Sun, May 13, 2018 at 9:43 AM, ASM Jonayat [email protected] wrote:
Just wanted to know if you have the reference for matminer now.
Regards,
JonayatOn Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/68d617fd-9410-45be-8f20-3c74394e53ee%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
Thanks, I will add the proper citation in my work. Should be out by July for publication.
···
On Friday, May 25, 2018 at 2:52:27 PM UTC-4, Anubhav Jain wrote:
Hi,
The matminer paper is now published, please see:
Thanks for your support!
Anubhav
On Monday, May 14, 2018 at 10:18:48 AM UTC-7, Anubhav Jain wrote:
Hi,
Thanks for checking up!
We have a paper that is accepted but not yet published in Computational Materials Science. I will update the thread again when the paper is actually published / available online.
Best,
Anubhav
On Sun, May 13, 2018 at 9:43 AM, ASM Jonayat [email protected] wrote:
Just wanted to know if you have the reference for matminer now.
Regards,
JonayatOn Friday, April 27, 2018 at 5:16:57 PM UTC-4, Anubhav Jain wrote:
Also, just an FYI that we have a paper on matminer currently in submission. If it ends up being published in time for your paper (not sure if that’ll be the case or not), you might want to refer to it.
Glad you’re finding matminer useful!
Anubhav
On Friday, April 27, 2018 at 11:46:48 AM UTC-7, Joseph Montoya wrote:
Great thanks for following up. I’ll still take a look to ensure it works with multiple featurizer. We’ll also start working on an example notebook.
Best,
Joey
On Apr 27, 2018, at 11:42 AM, ASM Jonayat [email protected] wrote:
Hello,
Thank you for the quick reply. I have figured it out for now. The issue was that I was using it inside the multiplefeaturizer. It works when I used it as a single featurizer. I am very excited about this function. I will be using it for my next publication.
It would be great to have a example on this on the website. For anyone who is working on feature generation from primary feature, this is a great tool.
Regards.
On Fri, Apr 27, 2018, 2:26 PM Joseph Montoya [email protected] wrote:
I’ll take a look shortly.
Best,
Joey
On Apr 27, 2018, at 11:22 AM, Anubhav Jain [email protected] wrote:
I am cc’ing Joey Montoya who should hopefully be able to answer (Joey, let me know if you need me to look at it)
I’ve also added a github issue to create a demo notebook for function featurizer
---------- Forwarded message ----------
From: ASM Jonayat [email protected]
Date: Fri, Apr 27, 2018 at 10:39 AM
Subject: Example for FunctionFeaturizer
To: matminer [email protected]Hello,
I am very interested to use the FunctionFeaturizer in matminer but was unable to run it. I have been writing my own matlab code to achieve this which worked (string operation was horribly slow) but I wanted to use a more general and transferable approach. Can anyone give me an example for this?
This was my attempt with a test dataset (df_test) -
from matminer.featurizers.function import FunctionFeaturizer
from matminer.featurizers.base import MultipleFeaturizerfeature_calculators = MultipleFeaturizer([FunctionFeaturizer(expressions=[“x*x”])])
feature_calculators.featurize_dataframe(df_test,col_id=[‘WF’]) #WF is one of the columns in df_test dataframe
``
and I get the following error.
--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-62-f1bbebfe2e41> in <module>() 4 feature_columns 5 feature_calculators.set_n_jobs(None) ----> 6 feature_labels = feature_calculators.feature_labels() ~/miniconda3/envs/matmine/lib/python3.6/site-packages/matminer/featurizers/base.py in feature_labels(self) 361 362 def feature_labels(self): --> 363 return sum([f.feature_labels() for f in self.featurizers], []) 364 365 def citations(self): TypeError: can only concatenate list (not "NoneType") to list
``
This coming from the feature_calculators.feature_
labels() but I can not figure out why.I uploaded the df_test datarame for convenience.
Regards.
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/8c140633-1ca9-4cf5-a9dc-24db7633eb61%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav
<df_test.pkl>
–
You received this message because you are subscribed to the Google Groups “matminer” group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/matminer/68d617fd-9410-45be-8f20-3c74394e53ee%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
–
Best,
Anubhav