Search
1000 results for “Data_Ranger”
-
The dollar-won exchange rate closed flat in the early 1,360 won range despite Federal Reserve Chair Jerome Powell's testimony, as risk appetite increased following a Middle East ceasefire and weak US consumer confidence data weighed on the dollar.
#YonhapInfomax #DollarWon #ExchangeRate #FederalReserve #ConsumerConfidence #TradingVolume #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=69155 -
The dollar-won exchange rate held in the upper 1,360 won range as risk aversion eased on hopes for de-escalation in the Middle East, with the greenback retreating after earlier gains; market focus remains on geopolitical developments and upcoming U.S. economic data.
#YonhapInfomax #DollarWon #ExchangeRate #MiddleEastRisk #Hezbollah #DollarIndex #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=68592 -
Nothing secret left
#Ukraine's military intelligence agency (HUR) gained access to sensitive data of #Russia's #strategic #aircraft manufacturer #Tupolev
Its bombers are widely used for long-range cruise missiles attacks
Over 4.4GB of internal data were obtained
"The significance of the data obtained cannot be overestimated," the source said. “Now, in fact, there is nothing secret left in Tupolev’s activities for #Ukrainian intelligence.”
-
"Ai2 tested DataDecide across a wide range of datasets and model sizes, using 10 benchmarks to evaluate how well small models predict large-scale performance. The findings aren’t earth-shattering, but they present useful takeaways for AI developers and researchers.
For one, Ai2 found that small models (around 150 million parameters) can predict large-scale outcomes with surprising accuracy. Some benchmarks reached over 80% decision accuracy using just 0.01% of the compute compared to billion-parameter models.
Since small-model experiments use less compute than other methods, developers don’t need to run full-scale tests just to predict outcomes. “The promise of this work is lower compute costs during training,” said Pijanowski.
Ai2 found that scaling laws didn’t outperform the simpler method of ranking datasets by small-model results. Scaling laws, a more sophisticated and more costly testing method, aim to predict how accuracy improves with model size. For now, “just stick with ablating things at one scale,” advised Magnusson.
The findings should give LLM devs pause for thought, Hunt said: “There are scaling laws that have been derived from empirical studies between data volume, compute resources and performance. Ai2’s research points out that we may want to revisit some of those assumptions.”"
https://thenewstack.io/new-tools-help-llm-developers-choose-better-pre-training-data/
-
"Ai2 tested DataDecide across a wide range of datasets and model sizes, using 10 benchmarks to evaluate how well small models predict large-scale performance. The findings aren’t earth-shattering, but they present useful takeaways for AI developers and researchers.
For one, Ai2 found that small models (around 150 million parameters) can predict large-scale outcomes with surprising accuracy. Some benchmarks reached over 80% decision accuracy using just 0.01% of the compute compared to billion-parameter models.
Since small-model experiments use less compute than other methods, developers don’t need to run full-scale tests just to predict outcomes. “The promise of this work is lower compute costs during training,” said Pijanowski.
Ai2 found that scaling laws didn’t outperform the simpler method of ranking datasets by small-model results. Scaling laws, a more sophisticated and more costly testing method, aim to predict how accuracy improves with model size. For now, “just stick with ablating things at one scale,” advised Magnusson.
The findings should give LLM devs pause for thought, Hunt said: “There are scaling laws that have been derived from empirical studies between data volume, compute resources and performance. Ai2’s research points out that we may want to revisit some of those assumptions.”"
https://thenewstack.io/new-tools-help-llm-developers-choose-better-pre-training-data/
-
"Ai2 tested DataDecide across a wide range of datasets and model sizes, using 10 benchmarks to evaluate how well small models predict large-scale performance. The findings aren’t earth-shattering, but they present useful takeaways for AI developers and researchers.
For one, Ai2 found that small models (around 150 million parameters) can predict large-scale outcomes with surprising accuracy. Some benchmarks reached over 80% decision accuracy using just 0.01% of the compute compared to billion-parameter models.
Since small-model experiments use less compute than other methods, developers don’t need to run full-scale tests just to predict outcomes. “The promise of this work is lower compute costs during training,” said Pijanowski.
Ai2 found that scaling laws didn’t outperform the simpler method of ranking datasets by small-model results. Scaling laws, a more sophisticated and more costly testing method, aim to predict how accuracy improves with model size. For now, “just stick with ablating things at one scale,” advised Magnusson.
The findings should give LLM devs pause for thought, Hunt said: “There are scaling laws that have been derived from empirical studies between data volume, compute resources and performance. Ai2’s research points out that we may want to revisit some of those assumptions.”"
https://thenewstack.io/new-tools-help-llm-developers-choose-better-pre-training-data/
-
"Ai2 tested DataDecide across a wide range of datasets and model sizes, using 10 benchmarks to evaluate how well small models predict large-scale performance. The findings aren’t earth-shattering, but they present useful takeaways for AI developers and researchers.
For one, Ai2 found that small models (around 150 million parameters) can predict large-scale outcomes with surprising accuracy. Some benchmarks reached over 80% decision accuracy using just 0.01% of the compute compared to billion-parameter models.
Since small-model experiments use less compute than other methods, developers don’t need to run full-scale tests just to predict outcomes. “The promise of this work is lower compute costs during training,” said Pijanowski.
Ai2 found that scaling laws didn’t outperform the simpler method of ranking datasets by small-model results. Scaling laws, a more sophisticated and more costly testing method, aim to predict how accuracy improves with model size. For now, “just stick with ablating things at one scale,” advised Magnusson.
The findings should give LLM devs pause for thought, Hunt said: “There are scaling laws that have been derived from empirical studies between data volume, compute resources and performance. Ai2’s research points out that we may want to revisit some of those assumptions.”"
https://thenewstack.io/new-tools-help-llm-developers-choose-better-pre-training-data/
-
"Ai2 tested DataDecide across a wide range of datasets and model sizes, using 10 benchmarks to evaluate how well small models predict large-scale performance. The findings aren’t earth-shattering, but they present useful takeaways for AI developers and researchers.
For one, Ai2 found that small models (around 150 million parameters) can predict large-scale outcomes with surprising accuracy. Some benchmarks reached over 80% decision accuracy using just 0.01% of the compute compared to billion-parameter models.
Since small-model experiments use less compute than other methods, developers don’t need to run full-scale tests just to predict outcomes. “The promise of this work is lower compute costs during training,” said Pijanowski.
Ai2 found that scaling laws didn’t outperform the simpler method of ranking datasets by small-model results. Scaling laws, a more sophisticated and more costly testing method, aim to predict how accuracy improves with model size. For now, “just stick with ablating things at one scale,” advised Magnusson.
The findings should give LLM devs pause for thought, Hunt said: “There are scaling laws that have been derived from empirical studies between data volume, compute resources and performance. Ai2’s research points out that we may want to revisit some of those assumptions.”"
https://thenewstack.io/new-tools-help-llm-developers-choose-better-pre-training-data/
-
Get Ready for the New Time Data Type – Summer ‘25 Flow Goodness
Salesforce Flow is constantly evolving, bringing us enhancements that make our lives as admins, developers, and business users much easier. The Summer ‘25 release is described as a big one, packed with substantial updates and quality-of-life improvements. Among these exciting additions is a feature many have been waiting for: native support for the Time data type in Flow.
What is the Time Data Type and Why is it Important?
The new Time data type is specifically designed for situations where the time of day matters, but the date does not. Previously, handling time-specific data in Flow without including the date could be complex. Summer ’25 changes that, allowing you to process data focused purely on time, down to the millisecond.
This capability is incredibly handy for a variety of use cases:
- Managing communication times, such as determining when to send an email.
- Checking if actions occur within specific business hours.
- Creating flows to send reminders based on a time before an event, like an email reminder 30 minutes before a meeting.
Where Can You Use the Time Data Type in Flow?
The Time data type is available across a wide range of Flow features, providing flexibility in how you build your automations. You can use Time fields and resources in:
- Various Flow elements, including Action, Assignment, Collection Filter, Collection Sort, Create Records, Delete Records, Decision, Get Records, Subflow, Transform, Update Records, and Wait for Conditions.
- Formula builder and expression builder.
- Resources such as variables and constants.
- As input and output for invocable actions.
When working with time values, you should use the
hh:mm:ss.SSS AM/PMformat, though including seconds or milliseconds is optional. For instance,9:00 AM,5:30:05 PM, and14:45:53.650 PMare all valid time values.New and Improved Time Functions
To complement the new data type, Salesforce Flow also introduces or enhances formula functions specifically for working with time. In the formula editor, you can now effectively use functions such as
HOUR(),MINUTE(),SECOND(),MILLISECOND(),TIMENOW(), andTIMEVALUE(). These functions empower you to perform calculations and make decisions based on time data within your flows. Previously, extracting and manipulating time in Date/Time fields was very difficult, and it involved parsing text values that contained this information.Important Considerations
- The Time data type is currently not supported in the offline flows available on the Salesforce Mobile app.
- This change applies to flows running on API version 64.0 or later. If you have existing flows created with API version 63.0 or earlier that use custom fields of the time data type, they will continue to work as before. However, to leverage the full functionality of the updated time data type in those flows, you’ll need to edit them and save them as a new version configured to run on API version 64.0.
Random Number Generation
One benefit of the new time-related capabilities is that you can use the new functions to generate random numbers. There is no random number generator function available in flow. Previously, I extracted the seconds out of a Date/Time value to generate a random number, now I can generate one using the Milliseconds.
🚨 Use Case 👇🏼Select multiple leads on a data table to add them to a prize drawing. Generate a random number and determine the winner. Email the winner to communicate the prize they won.
For this use case I leveraged many new flow functionalities.
Let’s get right to the build.
Build the Screen Action Autolaunched Flow
The selected leads can span over several screens in the data table, when the user is completing their selection. I decided to use an autolaunched flow to compile a CSV list of lead names which will be shown under the data table, as the user is completing their selection.
For that I build an autolaunched flow. Follow these instructions to build yours:
- Start an autolaunched flow.
- Create a Lead Collection Record Variable and make it available for input.
- Create a Name CSV Text Variable and make it available for output.
- Use the transform element to extract a text collection variable of names (first names) out of the lead collection record variable (not required, I wanted to use this new feature).
- Loop the names collection text variable.
- Add an assignment to add the current name text, and then a comma and a space character to the Name CSV Text Variable.
- Outside the loop add another assignment to assign a new value to the Name CSV Text Variable. This new value will be the accumulated names in csv format with the last comma and the space character removed. Use a formula resource to compute the value. The formula is:
LEFT({!NameCSVTextVar},LEN({!NameCSVTextVar})-2) - Debug, save and activate the flow.
Build the Screen Flow
Follow these instruction to build your flow:
- Start a screen flow.
- Get the leads in the org where the email is no null (limit the get to 2,000 records not to hit limits).
- Add a screen. Place a data table on the screen showing the leads, and allow for multi selection. Add a screen action to the screen and point it to the autolaunched flow you created above. Pass the Lead Data Table Selected Rows to the screen action autolaunched flow as input.
- Assign the count to a Count Number Variable (no decimals). Also assign the winner number to a Winner Count Variable. This is to ensure that the number does not change in debug (I don’t think it will change in production execution). You will need a formula resource to determine the winner. Here is what this formula does: Generate a number between 1 and 1,000 using the milliseconds value of the time of the execution, and prorate that using the number of leads the user selected to determine the winning number. Assign the following formula value to the Winner Count Variable:
ROUND(((MILLISECOND(TIMENOW())+1)*{!CountLeadsVar}/1000),0)+1 - Loop the Lead Data Table Selected Rows and assign a value incremented by 1 to a counter variable in every iteration (CounterVar Add 1).
- Check Via a decision whether the winning number is equal to the counter variable.
- If the winner is determined assign the Lead Record to the Winning Lead Record Variable, and exit loop. If not, keep looping.
- Outside the loop send the email to the email address of the winning lead and congratulate them. I built my email template with inside the brand new email action element for this one (Summer ’25).
- Debug, save and activate the flow.
Please note that, I tried conditionally running the screen action only after the user selects the first data table row, but that functionality (Summer ’25) does not seem to work properly in preview. I have a ticket open with Salesforce to determine whether that is a bug.
If you want to see the flow in action, watch this video.
Conclusion
The introduction of the Time data type is a significant step forward for Flow, enabling more precise and efficient time-based automation. It’s one of the many high-impact features and quality-of-life improvements packed into the Summer ’25 release that are bound to make your job easier.
Ready to give it a spin? Don’t forget to sign up for a pre-release org to test out this and other new features! You can also find more details in the Summer ’25 release notes.
Explore related content:
Salesforce Summer ’25 Preview: Major Flow Changes to Watch For
Time Zone and Time Operations in Flow
Supercharge Your Approvals with Salesforce Flow Approval Processes
#AutolaunchedFlow #DecisionElement #GetRecords #Salesforce #SalesforceAdmins #SalesforceDevelopers #SalesforceTutorials #Summer25 #Time #TimeDataType
-
Get Ready for the New Time Data Type – Summer ‘25 Flow Goodness
Salesforce Flow is constantly evolving, bringing us enhancements that make our lives as admins, developers, and business users much easier. The Summer ‘25 release is described as a big one, packed with substantial updates and quality-of-life improvements. Among these exciting additions is a feature many have been waiting for: native support for the Time data type in Flow.
What is the Time Data Type and Why is it Important?
The new Time data type is specifically designed for situations where the time of day matters, but the date does not. Previously, handling time-specific data in Flow without including the date could be complex. Summer ’25 changes that, allowing you to process data focused purely on time, down to the millisecond.
This capability is incredibly handy for a variety of use cases:
- Managing communication times, such as determining when to send an email.
- Checking if actions occur within specific business hours.
- Creating flows to send reminders based on a time before an event, like an email reminder 30 minutes before a meeting.
Where Can You Use the Time Data Type in Flow?
The Time data type is available across a wide range of Flow features, providing flexibility in how you build your automations. You can use Time fields and resources in:
- Various Flow elements, including Action, Assignment, Collection Filter, Collection Sort, Create Records, Delete Records, Decision, Get Records, Subflow, Transform, Update Records, and Wait for Conditions.
- Formula builder and expression builder.
- Resources such as variables and constants.
- As input and output for invocable actions.
When working with time values, you should use the
hh:mm:ss.SSS AM/PMformat, though including seconds or milliseconds is optional. For instance,9:00 AM,5:30:05 PM, and14:45:53.650 PMare all valid time values.New and Improved Time Functions
To complement the new data type, Salesforce Flow also introduces or enhances formula functions specifically for working with time. In the formula editor, you can now effectively use functions such as
HOUR(),MINUTE(),SECOND(),MILLISECOND(),TIMENOW(), andTIMEVALUE(). These functions empower you to perform calculations and make decisions based on time data within your flows. Previously, extracting and manipulating time in Date/Time fields was very difficult, and it involved parsing text values that contained this information.Important Considerations
- The Time data type is currently not supported in the offline flows available on the Salesforce Mobile app.
- This change applies to flows running on API version 64.0 or later. If you have existing flows created with API version 63.0 or earlier that use custom fields of the time data type, they will continue to work as before. However, to leverage the full functionality of the updated time data type in those flows, you’ll need to edit them and save them as a new version configured to run on API version 64.0.
Random Number Generation
One benefit of the new time-related capabilities is that you can use the new functions to generate random numbers. There is no random number generator function available in flow. Previously, I extracted the seconds out of a Date/Time value to generate a random number, now I can generate one using the Milliseconds.
🚨 Use Case 👇🏼Select multiple leads on a data table to add them to a prize drawing. Generate a random number and determine the winner. Email the winner to communicate the prize they won.
For this use case I leveraged many new flow functionalities.
Let’s get right to the build.
Build the Screen Action Autolaunched Flow
The selected leads can span over several screens in the data table, when the user is completing their selection. I decided to use an autolaunched flow to compile a CSV list of lead names which will be shown under the data table, as the user is completing their selection.
For that I build an autolaunched flow. Follow these instructions to build yours:
- Start an autolaunched flow.
- Create a Lead Collection Record Variable and make it available for input.
- Create a Name CSV Text Variable and make it available for output.
- Use the transform element to extract a text collection variable of names (first names) out of the lead collection record variable (not required, I wanted to use this new feature).
- Loop the names collection text variable.
- Add an assignment to add the current name text, and then a comma and a space character to the Name CSV Text Variable.
- Outside the loop add another assignment to assign a new value to the Name CSV Text Variable. This new value will be the accumulated names in csv format with the last comma and the space character removed. Use a formula resource to compute the value. The formula is:
LEFT({!NameCSVTextVar},LEN({!NameCSVTextVar})-2) - Debug, save and activate the flow.
Build the Screen Flow
Follow these instruction to build your flow:
- Start a screen flow.
- Get the leads in the org where the email is no null (limit the get to 2,000 records not to hit limits).
- Add a screen. Place a data table on the screen showing the leads, and allow for multi selection. Add a screen action to the screen and point it to the autolaunched flow you created above. Pass the Lead Data Table Selected Rows to the screen action autolaunched flow as input.
- Assign the count to a Count Number Variable (no decimals). Also assign the winner number to a Winner Count Variable. This is to ensure that the number does not change in debug (I don’t think it will change in production execution). You will need a formula resource to determine the winner. Here is what this formula does: Generate a number between 1 and 1,000 using the milliseconds value of the time of the execution, and prorate that using the number of leads the user selected to determine the winning number. Assign the following formula value to the Winner Count Variable:
ROUND(((MILLISECOND(TIMENOW())+1)*{!CountLeadsVar}/1000),0)+1 - Loop the Lead Data Table Selected Rows and assign a value incremented by 1 to a counter variable in every iteration (CounterVar Add 1).
- Check Via a decision whether the winning number is equal to the counter variable.
- If the winner is determined assign the Lead Record to the Winning Lead Record Variable, and exit loop. If not, keep looping.
- Outside the loop send the email to the email address of the winning lead and congratulate them. I built my email template with inside the brand new email action element for this one (Summer ’25).
- Debug, save and activate the flow.
Please note that, I tried conditionally running the screen action only after the user selects the first data table row, but that functionality (Summer ’25) does not seem to work properly in preview. I have a ticket open with Salesforce to determine whether that is a bug.
If you want to see the flow in action, watch this video.
Conclusion
The introduction of the Time data type is a significant step forward for Flow, enabling more precise and efficient time-based automation. It’s one of the many high-impact features and quality-of-life improvements packed into the Summer ’25 release that are bound to make your job easier.
Ready to give it a spin? Don’t forget to sign up for a pre-release org to test out this and other new features! You can also find more details in the Summer ’25 release notes.
Explore related content:
Salesforce Summer ’25 Preview: Major Flow Changes to Watch For
Time Zone and Time Operations in Flow
Supercharge Your Approvals with Salesforce Flow Approval Processes
#AutolaunchedFlow #DecisionElement #GetRecords #Salesforce #SalesforceAdmins #SalesforceDevelopers #SalesforceTutorials #Summer25 #Time #TimeDataType
-
GHGA Retreat – Day 1 at Kloster Schöntal kicked off with exciting discussions on a wide range of topics! From data portals to international connectivity (#FEGA, #ELIXIR, #EOSC), legal challenges, patient communication, and training plans, GHGA 2 is taking shape across all fronts. Perspectives from data producers and data hubs, along with workflow use cases, rounded off a fantastic first session. #GHGA #GHGAretreat #Genomics #OpenScience #Collaboration
-
GHGA Retreat – Day 1 at Kloster Schöntal kicked off with exciting discussions on a wide range of topics! From data portals to international connectivity (#FEGA, #ELIXIR, #EOSC), legal challenges, patient communication, and training plans, GHGA 2 is taking shape across all fronts. Perspectives from data producers and data hubs, along with workflow use cases, rounded off a fantastic first session. #GHGA #GHGAretreat #Genomics #OpenScience #Collaboration
-
GHGA Retreat – Day 1 at Kloster Schöntal kicked off with exciting discussions on a wide range of topics! From data portals to international connectivity (#FEGA, #ELIXIR, #EOSC), legal challenges, patient communication, and training plans, GHGA 2 is taking shape across all fronts. Perspectives from data producers and data hubs, along with workflow use cases, rounded off a fantastic first session. #GHGA #GHGAretreat #Genomics #OpenScience #Collaboration
-
GHGA Retreat – Day 1 at Kloster Schöntal kicked off with exciting discussions on a wide range of topics! From data portals to international connectivity (#FEGA, #ELIXIR, #EOSC), legal challenges, patient communication, and training plans, GHGA 2 is taking shape across all fronts. Perspectives from data producers and data hubs, along with workflow use cases, rounded off a fantastic first session. #GHGA #GHGAretreat #Genomics #OpenScience #Collaboration
-
GHGA Retreat – Day 1 at Kloster Schöntal kicked off with exciting discussions on a wide range of topics! From data portals to international connectivity (#FEGA, #ELIXIR, #EOSC), legal challenges, patient communication, and training plans, GHGA 2 is taking shape across all fronts. Perspectives from data producers and data hubs, along with workflow use cases, rounded off a fantastic first session. #GHGA #GHGAretreat #Genomics #OpenScience #Collaboration
-
🔍 New proposal: A vocabulary for opting out from AI training & text/data mining.
Based on interaction with a broad range of stakeholders, this proposal aims to give creators and other rightholders more control over how their works are used for AI training through practical, machine-readable standards.
📄 Full paper & vocabulary: https://openfuture.eu/publication/a-vocabulary-for-opting-out-of-ai-training-and-other-forms-of-tdm/
#ParadoxOfOpen #AITraining -
Seoul forex dealers anticipate USD/KRW to remain heavy in 1,440 won range, with mixed signals from US job data and Fed comments influencing market sentiment
#YonhapInfomax #USDKRWExchangeRate #SeoulForexMarket #DollarIndex #FederalReserve #NonFarmPayroll #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=53546 -
Won strengthens against dollar, entering 1,450 range amid global greenback weakness, as markets digest U.S. President's congressional address and await key employment data
#YonhapInfomax #DollarWonExchangeRate #GlobalDollarWeakness #TrumpCongressAddress #USEmploymentData #KoreanWonStrength #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=53066 -
U.S. Treasury yields surge as disappointing economic data and new COVID strain reports from China fuel safe-haven demand, pushing 10-year yield to low 4.4% range.
#YonhapInfomax #USTreasuryYields #EconomicIndicators #ChinaCOVID #FederalReserve #MarketSentiment #Economics #FinancialMarkets #Banking #Securities #Bonds #StockMarket
https://en.infomaxai.com/news/articleView.html?idxno=51742 -
Hi folks, my library is hiring a data librarian for the health sciences
Appointment will be at the Librarian 1 to 3 ranks, with a starting salary range of $66,000 - $90,000. In-person required at least 3 days a week
Posting with salary range on the med library website: https://library.medicine.yale.edu/about/job-opportunities/data-librarian-2025Posting without salary range on the HR website: https://sjobs.brassring.com/TGnewUI/Search/Home/Home?partnerid=25053&siteid=5248#jobDetails=1594109_5248
I'm not part of the hiring committee and am happy to answer questions
-
In Bayesian inference, a credible interval is a range of values within which a parameter lies with a certain probability, given the observed data and prior beliefs. The image of this post (based on this Wikipedia image: https://en.wikipedia.org/wiki/Credible_interval#/media/File:Highest_posterior_density_interval.svg) represents a 90% highest-density credible interval of a posterior probability distribution.
More details: http://eepurl.com/gH6myT
#statistical #datasciencecourse #datascience #rprogramming #datastructure
-
Open sourcerers say suspected #xz-style attacks continue to target #maintainers
#SocialEngineering patterns spotted across range of popular projects
Higher-ups at the #OpenJS Foundation and #OpenSource Security Foundation (#OpenSSF) believe the attempt to plant a #backdoor into #Linux's xz data compression library "may not be an isolated incident" given their recent observations.
https://www.theregister.com/2024/04/16/xz_style_attacks_continue/ -
Looking forward to DataConnect23 (https://dataconnect.api.gov.uk/) next week - a huge number of interesting data events around a range of important topics.
Hope to find out what others working on data and with data across government have to share.
We’ve written about some of the projects we’ve been proud to support on our project page and recent blog posts. See: https://www.epimorphics.com/blog and https://www.epimorphics.com/projects/ for more.
-
Pantera Capital seeks $1B for a new crypto fund: Report - The Pantera Fund V will invest in a range of blockchain-based assets and... - https://cointelegraph.com/news/pantera-capital-seeks-one-billion-raise-for-new-crypto-fund #panteracryptofundraise #andreessenhorowitz #panteracapital #panteracrypto #panteravfund #paradigm #funding #gamefi #data #aum #ai
-
State-level percentage declines in preschool enrollment from 2019 to 2021 ranged from 3.6 points in Indiana to 19.0 points in New Hampshire. California saw among the largest drops (13.9 points).
https://www.census.gov/library/stories/2023/08/preschool-enrollment.html
Image: Figure from U.S. Census Bureau, map showing percentage change in preschool enrollment by state, data available at link above.
#Preschool #Education #EarlyLearning #Pandemic #COVID #COVID19 @edutooters
-
I also happen to really like the Marshmallow library for data validation. It can check types, for missing fields, for extraneous fields, and semantic checks on the values of fields (whether a string contains a valid email address, if numbers are in a specific range, etc.). It then prints out detailed error messages for the violations. #DataEngineering #cseducation #teaching #softwaredevelopment
-
I'm using the proprietary #mySugr Android app to sync the data from my glucometer (and I'd love to replace it with something open source). One of the app's features is giving points for data input. Enter blood sugar, you get points, enter carbs, points, enter insulin doses, points, tag, points. You get the idea.
Now, insulin doses are split into meal and correction doses. Both grant points separately.
Effectively, this means that if you repeatedly suffer from elevated blood sugars and need to correct them, you get more points than if you're well in range. Makes sense, right?
-
I updated my electric vehicle charging maps with the latest data. Surprisingly, the number of fast charging stations in Canada and the USA has doubled in the past 18 months.
With the charging density that now exists across Canada and the USA, you can road trip almost anywhere without any range anxiety.
https://canadianveggie.com/2024/06/29/rapid-growth-ev-charging-stations-in-canada-and-usa-analyzed/
#EV #mapping #dataNerd #GIS #electrifyEverything #transportation
-
Tip #603
Start fresh by deleting History and other browsing data in Vivaldi on Android.
It can be convenient to find a page you recently visited or autofill forms with your information, but it’s also good to get rid of unnecessary data every now and then. In Vivaldi on Android you can choose what to delete and for what time period.
To delete your browsing data:
- Open the History Panel and tap on the broom icon in the bottom right corner.
Alternatively, go to Settings > Privacy and Security > Delete browsing data. - Select the time range you want to delete data for.
- Select the data types you want to delete.
- Tap on “Delete data”.
- Open the History Panel and tap on the broom icon in the bottom right corner.
-
#Processing #Python #py5 #genuary #genuary31 #トゥートProcessing
# https://iamkate.com/data/12-bit-rainbow/
palette = (
'#817', '#a35', '#c66', '#e94',
'#ed0', '#9d5', '#4d8', '#2cb',
'#0bc', '#09c', '#36b', '#639'
)def setup():
size(800, 800)
no_stroke()
background(0)def draw():
xc = yc = 400
for i in range(6):
m = 1 - abs(cos(radians(frame_count / 2))) ** 5
r = 150 + 150 * m
a = radians(frame_count / 2 + 60 * i)
x = xc + r * cos(a)
y = yc + r * sin(a)
fill(palette[i])
circle(x, y, 150)
r = 300 - 150 * m
a = a + radians(30)
x = xc + r * cos(a)
y = yc + r * sin(a)
fill(palette[-1 -i])
circle(x, y, 150)