Using Dataiku
Sort by:
191 - 200 of
263
- Is there an easy way to avoid that python recipes change the data types from all integers to floats?
Last answer by Alex_CombessieHello,
For controlling the column types in pandas dataframes, you can deactivate the "infer_with_pandas" option in the get_dataframe() method on the input dataset(s). You can find more information on this documentation:
https://doc.dataiku.com/dss/latest/api/python/datasets.html#dataiku.Dataset.get_dataframe
Cheers,
Alex
Last answer by Alex_CombessieHello,
For controlling the column types in pandas dataframes, you can deactivate the "infer_with_pandas" option in the get_dataframe() method on the input dataset(s). You can find more information on this documentation:
https://doc.dataiku.com/dss/latest/api/python/datasets.html#dataiku.Dataset.get_dataframe
Cheers,
Alex
- Hi, I have a dataframe in my DSS workflow which I want to change and store in a non-csv file within a folder. Assume my dataframe is called df and for the example you can recreate is as follows df = p…Last answer by Alex_CombessieYes, a "DSS" folder is just a regular filesystem folder where you can store anything you want.Last answer by Alex_CombessieYes, a "DSS" folder is just a regular filesystem folder where you can store anything you want.
- To use the tensorflow in the custom python model, the code needs to provide the methods fit() and predict(), like SK-Learn. The code below is the code that I think I need to use. import tensorflow as …Last answer byLast answer by Alex_CombessieHello,
Interesting question. The issue here is that tensorflow models cannot be serialized through pickle as weight matrices are saved to external files. In theory you can build a wrapper around keras classifier (with tensorflow backend) to make it pickleable. It works by saving weight matrices to memory. You can have a look at https://pypi.python.org/pypi/keras-pickle-wrapper/1.0.3. This is untested, so let us know if it works.
Having said that, you can perfectly use tensorflow in a Python recipe or notebook, outside of the "Custom model" interface.
Cheers,
Alex - Good aftrenoon, I am currently programming in R through the Dataiku and I have a very special question, how can I write this two logical characters in the same phrase? When I am working just with & or…Last answer by
Last answer by Joachim_ZenticiHum, this is definitely an R question and not DSS related but it looks indeed like a correct R syntax. I don't know what problem you are actually experiencing (in particular, error or output you don't expect) but I can tell you that you can indeed combine those boolean operators. If you find it easier to separate the computation into several steps that you can check individually, don't hesitate. - I created new API service and new API endpoint (prediction (Function (R))) for my project. Dataiku has handy functions in "Code" and "Test" windows. In "Code" window user can expand and collapse R cod…Last answer by
Last answer by Clément_StenacHi,
Thanks for the suggestions, we'll add them for consideration ! We have a new improved UI coming up for "test queries" in the coming weeks. - Hi, we are trying to do a SELECT statement on a system table from a Python recipe. We also tried it using an SQL recipe. However, we are getting the following error from Dataiku: declare cursor statem…Last answer byLast answer by Clément_StenacHi,
Unfortunately, there isn't. You might be able to copy information from the system table to a regular table and then query the regular table.
(Note that Aster DB is not officially supported) - I am trying to make the code using the R studio now. And I got the question. Is it possible to copy paste the code from R to Dataiku? Or Dataikuusing another way of the packaging that R studio? Thank …Last answer byLast answer by Clément_StenacHi,
Mostly you can copy/paste your code. The only part that might differ is how you read data. In RStudio, you might be using "read.csv" whereas in DSS you would use "dkuReadDataset". But all processing code can be exactly the same. - hi, I know we can expand DATA_DIR in Dataiku to get the data directory, which is usually data_dir. However, I would need to get this is in a Python recipe? Is there a way to know the data directory?Solution bySolution by Clément_StenacHi,
Use os.environ["DIP_HOME"] - Hi, I try some Python code in recipe to read csv / excel file from html and local csv. The code link below: read from html link_down = link_share + '/download' df_proj_list = pandas.read_excel(link_do…Last answer by
Last answer by tmgiangActually I copy the code into my jupyter notebook and it still work well.
Back to your comment, I try to print the columns names after read, and it prints corrects columns's name.
But when I explore the data in table view, it is still missing the first row (the column name)
191 - 200 of
26320