Stuck waiting on Microsoft to grant you access to Python in Excel? Don't wait! Check out my video on how you can get started NOW: ruclips.net/video/0iV4FtpSriY/видео.html
Still new to Python and learning. Really thankful for your videos. But I do know that likes, comments, and shares will get this video in front of more people.
Great question! Assuming you can't correct the data (which is common), here are some strategies you can employ: 1 - Use an algorithm that can handle missing values. 2 - If only a small percentage of observations have missing data, remove the observations. 3 - Remove the feature with missing data. 4 - Find a “proxy feature” for the feature with missing data. 5 - Fill in the missing data (i.e., impute the missing data).
@@DaveOnData Thank you for such a complete response! I tried putting in a value that would flag them as not actual data -99999 for example, but I think that just throws the who analysis off. Next I will try removing the rows with null data, then after characterizing the clusters, I will see where the rows with null data might fit back in. sort of a post-hoc assignment to a cluster. Thanks again!
This was great. Thank you for posting this. I have a question - can we read and write data from local folder ? For eg: can I run pandas.read_csv() and read a csv file located at C:/User ?
Python in Excel cannot directly access the local file system, networks, or the Excel object model for security. However, you can use Power Query to source data from local files, databases, APIs, etc. For more details on Python in Excel security, check out the following Microsoft article: support.microsoft.com/en-us/office/data-security-and-python-in-excel-33cc88a4-4a87-485e-9ff9-f35958278327
Hey, great video. I have very little experience in this area and I got this running in an hour or so. I am trying to cluster customer sights(lat/long) in to sales territories. The issue is that with K-means I get all different size clusters. I have done some digging, and I want to try kmedoids clustering. Is it possible to use this method in Excel/Python? Thanks
A quick scan of the libraries included with Anaconda Python (which Python in Excel uses) showed that k-medoids doesn't appear to be available at this time. The main machine learning library provided with Python in Excel is scikit-learn which does not offer k-medoids. However, you could use a local Python installation and use a package like the following: python-kmedoids.readthedocs.io/en/latest/
Microsoft has not announced the date for Python in Excel General Availability. However, you don't need to wait to start building skills. Check out this video on my channel for more details: ruclips.net/video/0iV4FtpSriY/видео.html
Stuck waiting on Microsoft to grant you access to Python in Excel? Don't wait! Check out my video on how you can get started NOW: ruclips.net/video/0iV4FtpSriY/видео.html
Still new to Python and learning. Really thankful for your videos.
But I do know that likes, comments, and shares will get this video in front of more people.
Thank you so much for your continued support of the channel! It is much appreciated.
Awesome series David!
Thank you! I am glad you are enjoying the content. I'm on a mission to uplevel the quality of Python in Excel content on RUclips.
hey @daveondata Is there any way that we can get the excel file so we can do this ourselves?
@FelixLizama - I've added a GitHub repo link to the video description where you can get the Excel workbook file.
@@DaveOnData thanks!.. you are the best!
hey David, How do you handle data sets that are incomplete, for example some rows have no value for a specific column?
Great question! Assuming you can't correct the data (which is common), here are some strategies you can employ:
1 - Use an algorithm that can handle missing values.
2 - If only a small percentage of observations have missing data, remove the observations.
3 - Remove the feature with missing data.
4 - Find a “proxy feature” for the feature with missing data.
5 - Fill in the missing data (i.e., impute the missing data).
@@DaveOnData Thank you for such a complete response! I tried putting in a value that would flag them as not actual data -99999 for example, but I think that just throws the who analysis off. Next I will try removing the rows with null data, then after characterizing the clusters, I will see where the rows with null data might fit back in. sort of a post-hoc assignment to a cluster. Thanks again!
You are welcome. Happy data sleuthing!
This was great. Thank you for posting this.
I have a question - can we read and write data from local folder ? For eg: can I run pandas.read_csv() and read a csv file located at C:/User ?
Python in Excel cannot directly access the local file system, networks, or the Excel object model for security. However, you can use Power Query to source data from local files, databases, APIs, etc. For more details on Python in Excel security, check out the following Microsoft article:
support.microsoft.com/en-us/office/data-security-and-python-in-excel-33cc88a4-4a87-485e-9ff9-f35958278327
BTW - I will have an upcoming video demonstrating sourcing 300,000 rows from SQL Server into Python in Excel via Power Query. Stay tuned!
Hey, great video. I have very little experience in this area and I got this running in an hour or so. I am trying to cluster customer sights(lat/long) in to sales territories. The issue is that with K-means I get all different size clusters. I have done some digging, and I want to try kmedoids clustering. Is it possible to use this method in Excel/Python? Thanks
A quick scan of the libraries included with Anaconda Python (which Python in Excel uses) showed that k-medoids doesn't appear to be available at this time. The main machine learning library provided with Python in Excel is scikit-learn which does not offer k-medoids. However, you could use a local Python installation and use a package like the following: python-kmedoids.readthedocs.io/en/latest/
@@DaveOnData thank you, I will try to do a local implementation of Python.
When are they fully releasing python in excel for everyone
Microsoft has not announced the date for Python in Excel General Availability. However, you don't need to wait to start building skills. Check out this video on my channel for more details: ruclips.net/video/0iV4FtpSriY/видео.html