Hello. I have a 590MB csv file, and the program doesn't seem to be able to handle it (it seems it loads all the data into the RAM, thus horribly slowing the computer). I have tried to copy to data to a SQL SERVER database and use the SQL Table widget instead of the File widget, but I did not succeed to connect to the database (I get the following error: "(20009, b'DB-Lib error message 20009, severity 9:\Unable to connect: Adaptive Server is unavailable or does not exist (DESKTOP-RABJ6QO\MIKA_SQLEXPRESS) ')"). So is there a simple way to load a large csv?
You could try bz2 as I did (which reduced a 35Mb file --> 2.4Mb; 149k observations; 10 features). However, I found that Orange3 consistently choked on this quantum of data. I suspect O3 has a place for rapid visualisation, prototyping and testing; for large-scale production and deployment...not so much.
Hi, i have a question: how big data Orange can handle? I need to analyze a 6MB excel file, it seems displaying error “NULL” value, but the data is complete and filled all column.
Hi, I want to load data from an Excel-Sheet. My problem is, that there is a problem when i'm using timestamps. I'm using three rows as the header of the feature and the format ist like: 2018-11-05 00:00:00. Is this wrong? Thanks!
Datetime columns should be in the ISO format, namely 2018-11-05 00:00:00 as suggested above. Make sure that Excel doesn't try to be too smart and change the field type for you. Setting it to Text usually works for me. Then you can use 'time' in the second row to mark datetime variables.
I think Google changed some thing with their shared links. :( See if the link has edit access. If the issue persists, please report the bug to our Github repository.
I have a question. When I try to import CSV file with very small numbers, say 1 x 10-7, the software brings up that column as categorical instead of numeric. How do I get it to properly import as numeric?
You can set all of this directly in the file widget. You can have more than one meta column. As for multi-label classification - Orange currently doesn't support it.
The way you showed us to pull data from Google Sheet is not working now. To draw data from Sheets to Orange, you MUST use URL Shortener. You CAN NOT just paste the long address of your Google Sheets file like this.
This software is beyond useless. I spent an hour trying to make a line plot of a single column data set of a simple harmonic oscillator, which the software failed miserably at (just a white page), with nothing in documentation to even remotely explain what it could be, or if there any specific formatting that needs to be followed...stick to just plain python.
Very nice to spew hate for an open source software without even remotely trying to understand how it works. You could ask on our Discord about it and we would happily explain why you *cannot* make a line plot with a single column, but sure, go ahead and hate.
@@OrangeDataMining There is a difference between 'spewing hate' and being verbal about disliking software that doesn't have any documentation on specific data formation, and is competing against matlab, excel and mathematica, octave dozens of other data analysis systems in the most basic of functionality. I most certainly did try, I wanted to want to use the software as it is advertised, and spent more than significant time tearing about your iris data set and others, and reading through widget docs to find SOME kind of formatting rules, of which there is none. I even went under the assumption that every y value must have an accompanying x, generating and merging data sets (like in matlibplot) for this to fail miserably aswell. The fact it's opensource doesn't mean anything. However, in my frustration with orange software as much as with your frustration with my comment. I will admit that it's hardly helpful to anyone to assume my lack of understanding of said software sold to be 'painlessly easy' makes an entire software package useless, which clearly it does not. Commenting in passive aggressive 'woa is us we do this for free why are you so mean' is just as pointless. A documented help forum, rather than some chat room with who knows what kind of activity and arm chair experts ready to answer in beyond cryptic riddles, as it seems to be in most IRC/discord related help channels, is as exactly as helpful as my comment. I'll leave this comment here, and admit my previous one was more out of frustration in the lack of basic documentation for a basic function and dealing with it in a childish way. After a day or two, I'll have assumed you've read this response and I'll delete the original comment.
@@smoothoctopus I have no problem with you saying Orange doesn't meet your needs or that it needs additional documentation (in this you are absolutely right). But saying it is beyond useless when so many people are actively using and loving it, is quite amusing. You are also comparing it against software that we are in no way competing against, which means you wish to use Orange for something that is not meant to be used for. As for formatting, I don't know how much easier can it be than importing comma-separated or Excel files. The fact that it is open source means everything. It means we don't owe anyone anything. What we do, we do out of our passion for teaching data science, a lot of us in our spare time. Including the help channel which you so actively dislike. If you don't like what we offer, please, there are many software solutions out there for you to try. No one is forcing Orange upon you.
Orange better then Apple :)
I love the way she says 'coluumns' .................. [jees ...I'm so lonely *sigh]
awesome tool and awesome series.
Hello. I have a 590MB csv file, and the program doesn't seem to be able to handle it (it seems it loads all the data into the RAM, thus horribly slowing the computer). I have tried to copy to data to a SQL SERVER database and use the SQL Table widget instead of the File widget, but I did not succeed to connect to the database (I get the following error: "(20009, b'DB-Lib error message 20009, severity 9:\Unable to connect: Adaptive Server is unavailable or does not exist (DESKTOP-RABJ6QO\MIKA_SQLEXPRESS)
')"). So is there a simple way to load a large csv?
You could try bz2 as I did (which reduced a 35Mb file --> 2.4Mb; 149k observations; 10 features). However, I found that Orange3 consistently choked on this quantum of data. I suspect O3 has a place for rapid visualisation, prototyping and testing; for large-scale production and deployment...not so much.
Hi, i have a question: how big data Orange can handle? I need to analyze a 6MB excel file, it seems displaying error “NULL” value, but the data is complete and filled all column.
im not getting the dialogue box as in this video file pasting dialogue box appears, plz help
Why does my Orange read error the URL when I Reloaded? As a note I have set Anyone on the internet with this link can edit. Thanks
Why not hdf5 ?
Hi, I want to load data from an Excel-Sheet. My problem is, that there is a problem when i'm using timestamps. I'm using three rows as the header of the feature and the format ist like: 2018-11-05 00:00:00. Is this wrong? Thanks!
Datetime columns should be in the ISO format, namely 2018-11-05 00:00:00 as suggested above. Make sure that Excel doesn't try to be too smart and change the field type for you. Setting it to Text usually works for me. Then you can use 'time' in the second row to mark datetime variables.
Why does she say "Orange incorrectly assumes hair color are class variable"? What does it mean? AND WHY DOES SHE say "let's fix this"?
What is a "class attribute"?
File error when I reload file data from url?
notification appears "read error"
Please help me
I am also getting a similar error (Read error:
HTTP Error 404: Not Found
) when trying to load data from google spread sheet with a URL link
I think Google changed some thing with their shared links. :( See if the link has edit access. If the issue persists, please report the bug to our Github repository.
good video. do you have any video that show how to add new data from existed data in orange?
please what is the "meta attribute"?
An attribute not used for computation, normally (but necessarily) a string attribute.
@@OrangeDataMining cool, what is the exact difference between meta and classes?
@@frankoi7022 Metas are just labels, extra information. A class is a target variables, what you wish to predict.
@OrangeDataMining How to apply weight to the dataset
how to save data in my pc ctrl+s what after this cant save
on top right u will see file click on that SOMEWHERE U WILL SEE save
How about DEAP / Seed dataset?
I have a question. When I try to import CSV file with very small numbers, say 1 x 10-7, the software brings up that column as categorical instead of numeric. How do I get it to properly import as numeric?
Set the column type as numeric. Click on the column and select column type.
Thank you so much ! This video is very helpful for me !
:,
why can't I set 'target value' when creating the 3-row table. can I have multiple meta data columns and class columns? (guessing yes)
You can set all of this directly in the file widget. You can have more than one meta column. As for multi-label classification - Orange currently doesn't support it.
I have installed psycopg2 but the sql table widget is still disabled, are there more steps after installing psycopg2? Thx
Please see: github.com/biolab/orange3/issues/2892
How to import json format file
I'm having problems in inserting datasets. Can someone help me?
Can you please try to be a little more specific?
how to import data from oracle Database
How to establish database < -- > Orange connection?
With SQL Table widget. You need psycopg2 or pymssql modules as backend and then connect via the widget (enter credentials).
Is there possibility to work with nosql databases or it is just for relational databases?
Orange supports only postgres and mssql. For now.
how about if I use python script widget for nosql or hadoop(hive) database connection?
Surely you can try. :)
The way you showed us to pull data from Google Sheet is not working now. To draw data from Sheets to Orange, you MUST use URL Shortener. You CAN NOT just paste the long address of your Google Sheets file like this.
Use it to ya advantage im bent n focused
Wow I'd watch Ida read a phone book.
I wonder where she picked up her English, I hear British, USA, and German English inflections in her speech.
It is very commom to non english native speakers
she doesnt provide the file. this is stupid.
Of course we don't provide the file. The idea is you repeat the process described in the video.
This software is beyond useless. I spent an hour trying to make a line plot of a single column data set of a simple harmonic oscillator, which the software failed miserably at (just a white page), with nothing in documentation to even remotely explain what it could be, or if there any specific formatting that needs to be followed...stick to just plain python.
Very nice to spew hate for an open source software without even remotely trying to understand how it works. You could ask on our Discord about it and we would happily explain why you *cannot* make a line plot with a single column, but sure, go ahead and hate.
@@OrangeDataMining There is a difference between 'spewing hate' and being verbal about disliking software that doesn't have any documentation on specific data formation, and is competing against matlab, excel and mathematica, octave dozens of other data analysis systems in the most basic of functionality. I most certainly did try, I wanted to want to use the software as it is advertised, and spent more than significant time tearing about your iris data set and others, and reading through widget docs to find SOME kind of formatting rules, of which there is none. I even went under the assumption that every y value must have an accompanying x, generating and merging data sets (like in matlibplot) for this to fail miserably aswell. The fact it's opensource doesn't mean anything.
However, in my frustration with orange software as much as with your frustration with my comment. I will admit that it's hardly helpful to anyone to assume my lack of understanding of said software sold to be 'painlessly easy' makes an entire software package useless, which clearly it does not.
Commenting in passive aggressive 'woa is us we do this for free why are you so mean' is just as pointless. A documented help forum, rather than some chat room with who knows what kind of activity and arm chair experts ready to answer in beyond cryptic riddles, as it seems to be in most IRC/discord related help channels, is as exactly as helpful as my comment. I'll leave this comment here, and admit my previous one was more out of frustration in the lack of basic documentation for a basic function and dealing with it in a childish way. After a day or two, I'll have assumed you've read this response and I'll delete the original comment.
@@smoothoctopus I have no problem with you saying Orange doesn't meet your needs or that it needs additional documentation (in this you are absolutely right). But saying it is beyond useless when so many people are actively using and loving it, is quite amusing. You are also comparing it against software that we are in no way competing against, which means you wish to use Orange for something that is not meant to be used for.
As for formatting, I don't know how much easier can it be than importing comma-separated or Excel files.
The fact that it is open source means everything. It means we don't owe anyone anything. What we do, we do out of our passion for teaching data science, a lot of us in our spare time. Including the help channel which you so actively dislike. If you don't like what we offer, please, there are many software solutions out there for you to try. No one is forcing Orange upon you.
cancelled
Annoying orange
There are only two genders? I feel highly offended!