Search Results

Search found 727 results on 30 pages for 'evaluation'.

Page 10/30 | < Previous Page | 6 7 8 9 10 11 12 13 14 15 16 17  | Next Page >

  • Is it poor design to create objects that only execute code during the constructor?

    - by Curtix
    In my design I am using objects that evaluate a data record. The constructor is called with the data record and type of evaluation as parameters and then the constructor calls all of the object's code necessary to evaluate the record. This includes using the type of evaluation to find additional parameter-like data in a text file. There are in the neighborhood of 250 unique evaluation types that use the same or similar code and unique parameters coming from the text file. Some of these evaluations use different code so I benefit a lot from this model because I can use inheritance and polymorphism. Once the object is created there isn't any need to execute additional code on the object (at least for now) and it is used more like a struct; its kept on a list and 3 properties are used later. I think this design is the easiest to understand, code, and read. A logical alternative I guess would be using functions that return score structs, but you can't inherit from methods so it would make it kind of sloppy imo. I am using vb.net and these classes will be used in an asp.net web app as well as in a distributed app. thanks for your input

    Read the article

  • News about Oracle Documaker Enterprise Edition

    - by Susanne Hale
    Updates come from the Documaker front on two counts: Oracle Documaker Awarded XCelent Award for Best Functionality Celent has published a NEW report entitled Document Automation Solution Vendors for Insurers 2011. In the evaluation, Oracle received the XCelent award for Functionality, which recognizes solutions as the leader in this category of the evaluation. According to Celent, “Insurers need to address issues related to the creation and handling of all sorts of documents. Key issues in document creation are complexity and volume. Today, most document automation vendors provide an array of features to cope with the complexity and volume of documents insurers need to generate.” The report ranks ten solution providers on Technology, Functionality, Market Penetration, and Services. Each profile provides detailed information about the vendor and its document automation system, the professional services and support staff it offers, product features, insurance customers and reference feedback, its technology, implementation process, and pricing.  A summary of the report is available at Celent’s web site. Documaker User Group in Wisconsin Holds First Meeting Oracle Documaker users in Wisconsin made the first Documaker User Group meeting a great success, with representation from eight companies. On April 19, over 25 attendees got together to share information, best practices, experiences and concepts related to Documaker and enterprise document automation; they were also able to share feedback with Documaker product management. One insurer shared how they publish and deliver documents to both internal and external customers as quickly and cost effectively as possible, since providing point of sale documents to the sales force in real time is crucial to obtaining and maintaining the book of business. They outlined best practices that ensure consistent development and testing strategies processes are in place to maximize performance and reliability. And, they gave an overview of the supporting applications they developed to monitor and improve performance as well as monitor and track each transaction. Wisconsin User Group meeting photos are posted on the Oracle Insurance Facebook page http://www.facebook.com/OracleInsurance. The Wisconsin User Group will meet again on October 26. If you and other Documaker customers in your area are interested in setting up a user group in your area, please contact Susanne Hale ([email protected]), (703) 927-0863.

    Read the article

  • How relevant are Brainbench scores when evaluating candidates?

    - by Newtopian
    I've seen many companies using certification services such as Brainbench when evaluating candidates. Most times they use it as a secondary screen prior to interview or as a validation to choose between candidates. What is your experience with Brainbench scores? Did you try the tests yourself, and if so do you feel the score is meaningful enough to be used as part of a hiring process? Difficult choice. Consensus seems to be that BB cert are not very good as a certification. The biggest argument was around the fact that some of the questions are too precise to form a good evaluation. this view can probably be tempered somewhat but still, to hold someone's future solely on the results of this evaluation would be irresponsible. That said, I still think it is possible to use them properly to gain additional objective knowledge on a candidate's level of expertise provided the test is done in a controlled environment ensuring that all taking it stand on equal footing. Thus I went with the answer that best reflected this view keeping in mind that it is still just an hour long 50ish multiple choice question to evaluate skills and knowledge that take years to acquire. To be taken with a grain of salt ! In short, The tests have value but weather or not they are worth the money is another debate. Thanks all for your time.

    Read the article

  • How often is seq used in Haskell production code?

    - by Giorgio
    I have some experience writing small tools in Haskell and I find it very intuitive to use, especially for writing filters (using interact) that process their standard input and pipe it to standard output. Recently I tried to use one such filter on a file that was about 10 times larger than usual and I got a Stack space overflow error. After doing some reading (e.g. here and here) I have identified two guidelines to save stack space (experienced Haskellers, please correct me if I write something that is not correct): Avoid recursive function calls that are not tail-recursive (this is valid for all functional languages that support tail-call optimization). Introduce seq to force early evaluation of sub-expressions so that expressions do not grow to large before they are reduced (this is specific to Haskell, or at least to languages using lazy evaluation). After introducing five or six seq calls in my code my tool runs smoothly again (also on the larger data). However, I find the original code was a bit more readable. Since I am not an experienced Haskell programmer I wanted to ask if introducing seq in this way is a common practice, and how often one will normally see seq in Haskell production code. Or are there any techniques that allow to avoid using seq too often and still use little stack space?

    Read the article

  • F# for the C# Programmer

    - by mbcrump
    Are you a C# Programmer and can’t make it past a day without seeing or hearing someone mention F#?  Today, I’m going to walk you through your first F# application and give you a brief introduction to the language. Sit back this will only take about 20 minutes. Introduction Microsoft's F# programming language is a functional language for the .NET framework that was originally developed at Microsoft Research Cambridge by Don Syme. In October 2007, the senior vice president of the developer division at Microsoft announced that F# was being officially productized to become a fully supported .NET language and professional developers were hired to create a team of around ten people to build the product version. In September 2008, Microsoft released the first Community Technology Preview (CTP), an official beta release, of the F# distribution . In December 2008, Microsoft announced that the success of this CTP had encouraged them to escalate F# and it is now will now be shipped as one of the core languages in Visual Studio 2010 , alongside C++, C# 4.0 and VB. The F# programming language incorporates many state-of-the-art features from programming language research and ossifies them in an industrial strength implementation that promises to revolutionize interactive, parallel and concurrent programming. Advantages of F# F# is the world's first language to combine all of the following features: Type inference: types are inferred by the compiler and generic definitions are created automatically. Algebraic data types: a succinct way to represent trees. Pattern matching: a comprehensible and efficient way to dissect data structures. Active patterns: pattern matching over foreign data structures. Interactive sessions: as easy to use as Python and Mathematica. High performance JIT compilation to native code: as fast as C#. Rich data structures: lists and arrays built into the language with syntactic support. Functional programming: first-class functions and tail calls. Expressive static type system: finds bugs during compilation and provides machine-verified documentation. Sequence expressions: interrogate huge data sets efficiently. Asynchronous workflows: syntactic support for monadic style concurrent programming with cancellations. Industrial-strength IDE support: multithreaded debugging, and graphical throwback of inferred types and documentation. Commerce friendly design and a viable commercial market. Lets try a short program in C# then F# to understand the differences. Using C#: Create a variable and output the value to the console window: Sample Program. using System;   namespace ConsoleApplication9 {     class Program     {         static void Main(string[] args)         {             var a = 2;             Console.WriteLine(a);             Console.ReadLine();         }     } } A breeze right? 14 Lines of code. We could have condensed it a bit by removing the “using” statment and tossing the namespace. But this is the typical C# program. Using F#: Create a variable and output the value to the console window: To start, open Visual Studio 2010 or Visual Studio 2008. Note: If using VS2008, then please download the SDK first before getting started. If you are using VS2010 then you are already setup and ready to go. So, click File-> New Project –> Other Languages –> Visual F# –> Windows –> F# Application. You will get the screen below. Go ahead and enter a name and click OK. Now, you will notice that the Solution Explorer contains the following: Double click the Program.fs and enter the following information. Hit F5 and it should run successfully. Sample Program. open System let a = 2        Console.WriteLine a As Shown below: Hmm, what? F# did the same thing in 3 lines of code. Show me the interactive evaluation that I keep hearing about. The F# development environment for Visual Studio 2010 provides two different modes of execution for F# code: Batch compilation to a .NET executable or DLL. (This was accomplished above). Interactive evaluation. (Demo is below) The interactive session provides a > prompt, requires a double semicolon ;; identifier at the end of a code snippet to force evaluation, and returns the names (if any) and types of resulting definitions and values. To access the F# prompt, in VS2010 Goto View –> Other Window then F# Interactive. Once you have the interactive window type in the following expression: 2+3;; as shown in the screenshot below: I hope this guide helps you get started with the language, please check out the following books for further information. F# Books for further reading   Foundations of F# Author: Robert Pickering An introduction to functional programming with F#. Including many samples, this book walks through the features of the F# language and libraries, and covers many of the .NET Framework features which can be leveraged with F#.       Functional Programming for the Real World: With Examples in F# and C# Authors: Tomas Petricek and Jon Skeet An introduction to functional programming for existing C# developers written by Tomas Petricek and Jon Skeet. This book explains the core principles using both C# and F#, shows how to use functional ideas when designing .NET applications and presents practical examples such as design of domain specific language, development of multi-core applications and programming of reactive applications.

    Read the article

  • JavaOne 2011: Content review process and Tips for submissions

    - by arungupta
    The Technical Sessions, Birds of Feather, Panels, and Hands-on labs (basically all the content delivered at JavaOne) forms the backbone of the conference. At this year's JavaOne conference you'll have access to the rock star speakers, the ability to engage with luminaries in the hallways, and have beer (or 2) with community peers in designated areas. Even though the conference is Oct 2-6, 2011, and will be bigger and better than last year's conference, the Call for Paper submission and review/selection evaluation started much earlier.In previous years, I've participated in the review process and this year I was honored to serve as co-lead for the "Enterprise Service Architecture and Cloud" track with Ludovic Champenois. We had a stellar review team with an equal mix of Oracle and external community reviewers. The review process is very overwhelming with the reviewers going through multiple voting iterations on each submission in order to ensure that the selected content is the BEST of the submitted lot. Our ultimate goal was to ensure that the content best represented the track, and most importantly would draw interest and excitement from attendees. As always, the number and quality of submissions were just superb, making for a truly challenging (and rewarding) experience for the reviewers. As co-lead I tried to ensure that I applied a fair and balanced process in the evaluation of content in my track. . Here are some key steps followed by all track leads: Vote on sessions - Each reviewer is required to vote on the sessions on a scale of 1-5 - and also provide a justifying comment. Create buckets - Divide the submissions into different buckets to ensure a fair representation of different topics within a track. This ensures that if a particular bucket got higher votes then the track is not exclusively skewed towards it. Top 7 - The review committee provides a list of the top 7 talks that can be used in the promotional material by the JavaOne team. Generally these talks are easy to identify and a consensus is reached upon them fairly quickly. First cut - Each track is allocated a total number of sessions (including panels), BoFs, and Hands-on labs that can be approved. The track leads then start creating the first cut of the approvals using the casted votes coupled with their prior experience in the subject matter. In our case, Ludo and I have been attending/speaking at JavaOne (and other popular Java-focused conferences) for double digit years. The Grind - The first cut is then refined and refined and refined using multiple selection criteria such as sorting on the bucket, speaker quality, topic popularity, cumulative vote total, and individual vote scale. The sessions that don't make the cut are reviewed again as well to ensure if they need to replace one of the selected one as a potential alternate. I would like to thank the entire Java community for all the submissions and many thanks to the reviewers who spent countless hours reading each abstract, voting on them, and helping us refine the list. I think approximately 3-4 hours cumulative were spent on each submission to reach an evaluation, specifically the border line cases. We gave our recommendations to the JavaOne Program Committee Chairperson (Sharat Chander) and accept/decline notifications should show up in submitter inboxes in the next few weeks. Here are some points to keep in mind when submitting a session to JavaOne next time: JavaOne is a technology-focused conference so any product, marketing or seemingly marketish talk are put at the bottom of the list.Oracle Open World and Oracle Develop are better options for submitting product specific talks. Make your title catchy. Remember the attendees are more likely to read the abstract if they like the title. We try our best to recategorize the talk to a different track if it needs to but please ensure that you are filing in the right track to have all the right eyeballs looking at it. Also, it does not hurt marking an alternate track if your talk meets the criteria. Make sure to coordinate within your team before the submission - multiple sessions from the same team or company does not ensure that the best speaker is picked. In such case we rely upon your "google presence" and/or review committee's prior knowledge of the speaker. The reviewers may not know you or your product at all and you get 750 characters to pitch your idea. Make sure to use all of them, to the last 750th character. Make sure to read your abstract multiple times to ensure that you are giving all the relevant information ? Think through your presentation and see if you are leaving out any important aspects.Also look if the abstract has any redundant information that will not required by the reviewers. There are additional sections that allow you to share information about the speaker and the presentation summary. Use them to blow the horn about yourself and any other relevant details. Please don't say "call me at xxx-xxx-xxxx to find out the details" :-) The review committee enjoyed reviewing the submissions and we certainly hope you'll have a great time attending them. Happy JavaOne!

    Read the article

  • JavaOne 2011: Content review process and Tips for submissions

    - by arungupta
    The Technical Sessions, Birds of Feather, Panels, and Hands-on labs (basically all the content delivered at JavaOne) forms the backbone of the conference. At this year's JavaOne conference you'll have access to the rock star speakers, the ability to engage with luminaries in the hallways, and have beer (or 2) with community peers in designated areas. Even though the conference is Oct 2-6, 2011, and will be bigger and better than last year's conference, the Call for Paper submission and review/selection evaluation started much earlier.In previous years, I've participated in the review process and this year I was honored to serve as co-lead for the "Enterprise Service Architecture and Cloud" track with Ludovic Champenois. We had a stellar review team with an equal mix of Oracle and external community reviewers. The review process is very overwhelming with the reviewers going through multiple voting iterations on each submission in order to ensure that the selected content is the BEST of the submitted lot. Our ultimate goal was to ensure that the content best represented the track, and most importantly would draw interest and excitement from attendees. As always, the number and quality of submissions were just superb, making for a truly challenging (and rewarding) experience for the reviewers. As co-lead I tried to ensure that I applied a fair and balanced process in the evaluation of content in my track. . Here are some key steps followed by all track leads: Vote on sessions - Each reviewer is required to vote on the sessions on a scale of 1-5 - and also provide a justifying comment. Create buckets - Divide the submissions into different buckets to ensure a fair representation of different topics within a track. This ensures that if a particular bucket got higher votes then the track is not exclusively skewed towards it. Top 7 - The review committee provides a list of the top 7 talks that can be used in the promotional material by the JavaOne team. Generally these talks are easy to identify and a consensus is reached upon them fairly quickly. First cut - Each track is allocated a total number of sessions (including panels), BoFs, and Hands-on labs that can be approved. The track leads then start creating the first cut of the approvals using the casted votes coupled with their prior experience in the subject matter. In our case, Ludo and I have been attending/speaking at JavaOne (and other popular Java-focused conferences) for double digit years. The Grind - The first cut is then refined and refined and refined using multiple selection criteria such as sorting on the bucket, speaker quality, topic popularity, cumulative vote total, and individual vote scale. The sessions that don't make the cut are reviewed again as well to ensure if they need to replace one of the selected one as a potential alternate. I would like to thank the entire Java community for all the submissions and many thanks to the reviewers who spent countless hours reading each abstract, voting on them, and helping us refine the list. I think approximately 3-4 hours cumulative were spent on each submission to reach an evaluation, specifically the border line cases. We gave our recommendations to the JavaOne Program Committee Chairperson (Sharat Chander) and accept/decline notifications should show up in submitter inboxes in the next few weeks. Here are some points to keep in mind when submitting a session to JavaOne next time: JavaOne is a technology-focused conference so any product, marketing or seemingly marketish talk are put at the bottom of the list.Oracle Open World and Oracle Develop are better options for submitting product specific talks. Make your title catchy. Remember the attendees are more likely to read the abstract if they like the title. We try our best to recategorize the talk to a different track if it needs to but please ensure that you are filing in the right track to have all the right eyeballs looking at it. Also, it does not hurt marking an alternate track if your talk meets the criteria. Make sure to coordinate within your team before the submission - multiple sessions from the same team or company does not ensure that the best speaker is picked. In such case we rely upon your "google presence" and/or review committee's prior knowledge of the speaker. The reviewers may not know you or your product at all and you get 750 characters to pitch your idea. Make sure to use all of them, to the last 750th character. Make sure to read your abstract multiple times to ensure that you are giving all the relevant information ? Think through your presentation and see if you are leaving out any important aspects.Also look if the abstract has any redundant information that will not required by the reviewers. There are additional sections that allow you to share information about the speaker and the presentation summary. Use them to blow the horn about yourself and any other relevant details. Please don't say "call me at xxx-xxx-xxxx to find out the details" :-) The review committee enjoyed reviewing the submissions and we certainly hope you'll have a great time attending them. Happy JavaOne!

    Read the article

  • Best Method For Evaluating Existing Software or New Software

    How many of us have been faced with having to decide on an off-the-self or a custom built component, application, or solution to integrate in to an existing system or to be the core foundation of a new system? What is the best method for evaluating existing software or new software still in the design phase? One of the industry preferred methodologies to use is the Active Reviews for Intermediate Designs (ARID) evaluation process.  ARID is a hybrid mixture of the Active Design Review (ADR) methodology and the Architectural Tradeoff Analysis Method (ATAM). So what is ARID? ARD’s main goal is to ensure quality, detailed designs in software. One way in which it does this is by empowering reviewers by assigning generic open ended survey questions. This approach attempts to remove the possibility for allowing the standard answers such as “Yes” or “No”. The ADR process ignores the “Yes”/”No” questions due to the fact that they can be leading based on how the question is asked. Additionally these questions tend to receive less thought in comparison to more open ended questions. Common Active Design Review Questions What possible exceptions can occur in this component, application, or solution? How should exceptions be handled in this component, application, or solution? Where should exceptions be handled in this component, application, or solution? How should the component, application, or solution flow based on the design? What is the maximum execution time for every component, application, or solution? What environments can this component, application, or solution? What data dependencies does this component, application, or solution have? What kind of data does this component, application, or solution require? Ok, now I know what ARID is, how can I apply? Let’s imagine that your organization is going to purchase an off-the-shelf (OTS) solution for its customer-relationship management software. What process would we use to ensure that the correct purchase is made? If we use ARID, then we will have a series of 9 steps broken up by 2 phases in order to ensure that the correct OTS solution is purchases. Phase 1 Identify the Reviewers Prepare the Design Briefing Prepare the Seed Scenarios Prepare the Materials When identifying reviewers for a design it is preferred that they be pulled from a candidate pool comprised of developers that are going to implement the design. The believe is that developers actually implementing the design will have more a vested interest in ensuring that the design is correct prior to the start of code. Design debriefing consist of a summary of the design, examples of the design solving real world examples put in to use and should be no longer than two hours typically. The primary goal of this briefing is to adequately summarize the design so that the review members could actually implement the design. In the example of purchasing an OTS product I would attempt to review my briefing prior to its distribution with the review facilitator to ensure that nothing was excluded that should have not been. This practice will also allow me to test the length of the briefing to ensure that can be delivered in an appropriate about of time. Seed Scenarios are designed to illustrate conceptualized scenarios when applied with a set of sample data. These scenarios can then be used by the reviewers in the actual evaluation of the software, All materials needed for the evaluation should be prepared ahead of time so that they can be reviewed prior to and during the meeting. Materials Included: Presentation Seed Scenarios Review Agenda Phase 2 Present ARID Present Design Brainstorm and prioritize scenarios Apply scenarios Summarize Prior to the start of any ARID review meeting the Facilitator should define the remaining steps of ARID so that all the participants know exactly what they are doing prior to the start of the review process. Once the ARID rules have been laid out, then the lead designer presents an overview of the design which typically takes about two hours. During this time no questions about the design or rational are allowed to be asked by the review panel as a standard, but they are written down for use latter in the process. After the presentation the list of compiled questions is then summarized and sent back to the lead designer as areas that need to be addressed further. In the example of purchasing an OTS product issues could arise regarding security, the implementation needed or even if this is this the correct product to solve the needed solution. After the Design presentation a brainstorming and prioritize scenarios process begins by reducing the seed scenarios down to just the highest priority scenarios.  These will then be used to test the design for suitability. Once the selected scenarios have been defined the reviewers apply the examples provided in the presentation to the scenarios. The intended output of this process is to provide code or pseudo code that makes use of the examples provided while solving the selected seed scenarios. As a standard rule, the designers of the systems are not allowed to help the review board unless they all become stuck. When this occurs it is documented and along with the reason why the designer needed to help the review panel back on track. Once all of the scenarios have been completed the review facilitator reviews with the group issues that arise during the process. Then the reviewers will be polled as to efficacy of the review experience. References: Clements, Paul., Kazman, Rick., Klien, Mark. (2002). Evaluating Software Architectures: Methods and Case Studies Indianapolis, IN: Addison-Wesley

    Read the article

  • python- scipy optimization

    - by pear
    In scipy fmin_slsqp (Sequential Least Squares Quadratic Programming), I tried reading the code 'slsqp.py' provided with the scipy package, to find what are the criteria to get the exit_modes 0? I cannot find which statements in the code produce this exit mode? Please help me 'slsqp.py' code as follows, exit_modes = { -1 : "Gradient evaluation required (g & a)", 0 : "Optimization terminated successfully.", 1 : "Function evaluation required (f & c)", 2 : "More equality constraints than independent variables", 3 : "More than 3*n iterations in LSQ subproblem", 4 : "Inequality constraints incompatible", 5 : "Singular matrix E in LSQ subproblem", 6 : "Singular matrix C in LSQ subproblem", 7 : "Rank-deficient equality constraint subproblem HFTI", 8 : "Positive directional derivative for linesearch", 9 : "Iteration limit exceeded" } def fmin_slsqp( func, x0 , eqcons=[], f_eqcons=None, ieqcons=[], f_ieqcons=None, bounds = [], fprime = None, fprime_eqcons=None, fprime_ieqcons=None, args = (), iter = 100, acc = 1.0E-6, iprint = 1, full_output = 0, epsilon = _epsilon ): # Now do a lot of function wrapping # Wrap func feval, func = wrap_function(func, args) # Wrap fprime, if provided, or approx_fprime if not if fprime: geval, fprime = wrap_function(fprime,args) else: geval, fprime = wrap_function(approx_fprime,(func,epsilon)) if f_eqcons: # Equality constraints provided via f_eqcons ceval, f_eqcons = wrap_function(f_eqcons,args) if fprime_eqcons: # Wrap fprime_eqcons geval, fprime_eqcons = wrap_function(fprime_eqcons,args) else: # Wrap approx_jacobian geval, fprime_eqcons = wrap_function(approx_jacobian, (f_eqcons,epsilon)) else: # Equality constraints provided via eqcons[] eqcons_prime = [] for i in range(len(eqcons)): eqcons_prime.append(None) if eqcons[i]: # Wrap eqcons and eqcons_prime ceval, eqcons[i] = wrap_function(eqcons[i],args) geval, eqcons_prime[i] = wrap_function(approx_fprime, (eqcons[i],epsilon)) if f_ieqcons: # Inequality constraints provided via f_ieqcons ceval, f_ieqcons = wrap_function(f_ieqcons,args) if fprime_ieqcons: # Wrap fprime_ieqcons geval, fprime_ieqcons = wrap_function(fprime_ieqcons,args) else: # Wrap approx_jacobian geval, fprime_ieqcons = wrap_function(approx_jacobian, (f_ieqcons,epsilon)) else: # Inequality constraints provided via ieqcons[] ieqcons_prime = [] for i in range(len(ieqcons)): ieqcons_prime.append(None) if ieqcons[i]: # Wrap ieqcons and ieqcons_prime ceval, ieqcons[i] = wrap_function(ieqcons[i],args) geval, ieqcons_prime[i] = wrap_function(approx_fprime, (ieqcons[i],epsilon)) # Transform x0 into an array. x = asfarray(x0).flatten() # Set the parameters that SLSQP will need # meq = The number of equality constraints if f_eqcons: meq = len(f_eqcons(x)) else: meq = len(eqcons) if f_ieqcons: mieq = len(f_ieqcons(x)) else: mieq = len(ieqcons) # m = The total number of constraints m = meq + mieq # la = The number of constraints, or 1 if there are no constraints la = array([1,m]).max() # n = The number of independent variables n = len(x) # Define the workspaces for SLSQP n1 = n+1 mineq = m - meq + n1 + n1 len_w = (3*n1+m)*(n1+1)+(n1-meq+1)*(mineq+2) + 2*mineq+(n1+mineq)*(n1-meq) \ + 2*meq + n1 +(n+1)*n/2 + 2*m + 3*n + 3*n1 + 1 len_jw = mineq w = zeros(len_w) jw = zeros(len_jw) # Decompose bounds into xl and xu if len(bounds) == 0: bounds = [(-1.0E12, 1.0E12) for i in range(n)] elif len(bounds) != n: raise IndexError, \ 'SLSQP Error: If bounds is specified, len(bounds) == len(x0)' else: for i in range(len(bounds)): if bounds[i][0] > bounds[i][1]: raise ValueError, \ 'SLSQP Error: lb > ub in bounds[' + str(i) +'] ' + str(bounds[4]) xl = array( [ b[0] for b in bounds ] ) xu = array( [ b[1] for b in bounds ] ) # Initialize the iteration counter and the mode value mode = array(0,int) acc = array(acc,float) majiter = array(iter,int) majiter_prev = 0 # Print the header if iprint >= 2 if iprint >= 2: print "%5s %5s %16s %16s" % ("NIT","FC","OBJFUN","GNORM") while 1: if mode == 0 or mode == 1: # objective and constraint evaluation requird # Compute objective function fx = func(x) # Compute the constraints if f_eqcons: c_eq = f_eqcons(x) else: c_eq = array([ eqcons[i](x) for i in range(meq) ]) if f_ieqcons: c_ieq = f_ieqcons(x) else: c_ieq = array([ ieqcons[i](x) for i in range(len(ieqcons)) ]) # Now combine c_eq and c_ieq into a single matrix if m == 0: # no constraints c = zeros([la]) else: # constraints exist if meq > 0 and mieq == 0: # only equality constraints c = c_eq if meq == 0 and mieq > 0: # only inequality constraints c = c_ieq if meq > 0 and mieq > 0: # both equality and inequality constraints exist c = append(c_eq, c_ieq) if mode == 0 or mode == -1: # gradient evaluation required # Compute the derivatives of the objective function # For some reason SLSQP wants g dimensioned to n+1 g = append(fprime(x),0.0) # Compute the normals of the constraints if fprime_eqcons: a_eq = fprime_eqcons(x) else: a_eq = zeros([meq,n]) for i in range(meq): a_eq[i] = eqcons_prime[i](x) if fprime_ieqcons: a_ieq = fprime_ieqcons(x) else: a_ieq = zeros([mieq,n]) for i in range(mieq): a_ieq[i] = ieqcons_prime[i](x) # Now combine a_eq and a_ieq into a single a matrix if m == 0: # no constraints a = zeros([la,n]) elif meq > 0 and mieq == 0: # only equality constraints a = a_eq elif meq == 0 and mieq > 0: # only inequality constraints a = a_ieq elif meq > 0 and mieq > 0: # both equality and inequality constraints exist a = vstack((a_eq,a_ieq)) a = concatenate((a,zeros([la,1])),1) # Call SLSQP slsqp(m, meq, x, xl, xu, fx, c, g, a, acc, majiter, mode, w, jw) # Print the status of the current iterate if iprint > 2 and the # major iteration has incremented if iprint >= 2 and majiter > majiter_prev: print "%5i %5i % 16.6E % 16.6E" % (majiter,feval[0], fx,linalg.norm(g)) # If exit mode is not -1 or 1, slsqp has completed if abs(mode) != 1: break majiter_prev = int(majiter) # Optimization loop complete. Print status if requested if iprint >= 1: print exit_modes[int(mode)] + " (Exit mode " + str(mode) + ')' print " Current function value:", fx print " Iterations:", majiter print " Function evaluations:", feval[0] print " Gradient evaluations:", geval[0] if not full_output: return x else: return [list(x), float(fx), int(majiter), int(mode), exit_modes[int(mode)] ]

    Read the article

  • Why does Scala apply thunks automatically, sometimes?

    - by Anonymouse
    At just after 2:40 in ShadowofCatron's Scala Tutorial 3 video, it's pointed out that the parentheses following the name of a thunk are optional. "Buh?" said my functional programming brain, since the value of a function and the value it evaluates to when applied are completely different things. So I wrote the following to try this out. My thought process is described in the comments. object Main { var counter: Int = 10 def f(): Int = { counter = counter + 1; counter } def runThunk(t: () => Int): Int = { t() } def main(args: Array[String]): Unit = { val a = f() // I expect this to mean "apply f to no args" println(a) // and apparently it does val b = f // I expect this to mean "the value f", a function value println(b) // but it's the value it evaluates to when applied to no args println(b) // and the evaluation happens immediately, not in the call runThunk(b) // This is an error: it's not println doing something funny runThunk(f) // Not an error: seems to be val doing something funny } }   To be clear about the problem, this Scheme program (and the console dump which follows) shows what I expected the Scala program to do. (define counter (list 10)) (define f (lambda () (set-car! counter (+ (car counter) 1)) (car counter))) (define runThunk (lambda (t) (t))) (define main (lambda args (let ((a (f)) (b f)) (display a) (newline) (display b) (newline) (display b) (newline) (runThunk b) (runThunk f)))) > (main) 11 #<procedure:f> #<procedure:f> 13   After coming to this site to ask about this, I came across this answer which told me how to fix the above Scala program: val b = f _ // Hey Scala, I mean f, not f() But the underscore 'hint' is only needed sometimes. When I call runThunk(f), no hint is required. But when I 'alias' f to b with a val then apply it, it doesn't work: the evaluation happens in the val; and even lazy val works this way, so it's not the point of evaluation causing this behaviour.   That all leaves me with the question: Why does Scala sometimes automatically apply thunks when evaluating them? Is it, as I suspect, type inference? And if so, shouldn't a type system stay out of the language's semantics? Is this a good idea? Do Scala programmers apply thunks rather than refer to their values so much more often that making the parens optional is better overall? Examples written using Scala 2.8.0RC3, DrScheme 4.0.1 in R5RS.

    Read the article

  • How to make external Mathematica functions interruptible?

    - by Szabolcs
    I had an earlier question about integrating Mathematica with functions written in C++. This is a follow-up question: If the computation takes too long I'd like to be able to abort it using Evaluation Abort Evaluation. Which of the technologies suggested in the answers make it possible to have an interruptible C-based extension function? How can "interruptibility" be implemented on the C side? I need to make my function interruptible in a way which will corrupt neither it, nor the Mathematica kernel (i.e. it should be possible to call the function again from Mathematica after it has been interrupted)

    Read the article

  • Using Apache Velocity with StringBuilders/CharSequences

    - by mindas
    We are using Apache Velocity for dynamic templates. At the moment Velocity has following methods for evaluation/replacing: public static boolean evaluate(Context context, Writer writer, String logTag, Reader reader) public static boolean evaluate(Context context, Writer out, String logTag, String instring) We use these methods by providing StringWriter to write evaluation results. Our incoming data is coming in StringBuilder format so we use StringBuilder.toString and feed it as instring. The problem is that our templates are fairly large (can be megabytes, tens of Ms on rare cases), replacements occur very frequently and each replacement operation triples the amount of required memory (incoming data + StringBuilder.toString() which creates a new copy + outgoing data). I was wondering if there is a way to improve this. E.g. if I could find a way to provide a Reader and Writer on top of same StringBuilder instance that only uses extra memory for in/out differences, would that be a good approach? Has anybody done anything similar and could share any source for such a class? Or maybe there any better solutions to given problem?

    Read the article

  • C++ coverage tool that WORKS?!

    - by Poni
    Been searching for a good coverage tool for a while now. I'm with VC++ 2008. Tried already: NCover, PureCoverage, PartCover and a few others I can't remember their names. None works! Not even with a very basic console application that does almost nothing. Tried to get an evaluation copy from BullsEye and few more others - not only that they require you to ask for it (no automatic process), they don't even answer after you request AND after you ask what's going on, through the support mail. So the question is: Is there any tool, which I can immediately download an evaluation for it, that actually works on VC 2008 projects? Something that will get me going within 10 minutes? It's funny but it seems to be a very non-trivial request in the area of this kind of software.

    Read the article

  • Slides of my HOL on MySQL Cluster

    - by user13819847
    Hi!Thanks everyone who attended my hands-on lab on MySQL Cluster at MySQL Connect last Saturday.The following are the links for the slides, the HOL instructions, and the code examples.I'll try to summarize my HOL below.Aim of the HOL was to help attendees to familiarize with MySQL Cluster. In particular, by learning: the basics of MySQL Cluster Architecture the basics of MySQL Cluster Configuration and Administration how to start a new Cluster for evaluation purposes and how to connect to it We started by introducing MySQL Cluster. MySQL Cluster is a proven technology that today is successfully servicing the most performance-intensive workloads. MySQL Cluster is deployed across telecom networks and is powering mission-critical web applications. Without trading off use of commodity hardware, transactional consistency and use of complex queries, MySQL Cluster provides: Web Scalability (web-scale performance on both reads and writes) Carrier Grade Availability (99.999%) Developer Agility (freedom to use SQL or NoSQL access methods) MySQL Cluster implements: an Auto-Sharding, Multi-Master, Shared-nothing Architecture, where independent nodes can scale horizontally on commodity hardware with no shared disks, no shared memory, no single point of failure In the architecture of MySQL Cluster it is possible to find three types of nodes: management nodes: responsible for reading the configuration files, maintaining logs, and providing an interface to the administration of the entire cluster data nodes: where data and indexes are stored api nodes: provide the external connectivity (e.g. the NDB engine of the MySQL Server, APIs, Connectors) MySQL Cluster is recommended in the situations where: it is crucial to reduce service downtime, because this produces a heavy impact on business sharding the database to scale write performance higly impacts development of application (in MySQL Cluster the sharding is automatic and transparent to the application) there are real time needs there are unpredictable scalability demands it is important to have data-access flexibility (SQL & NoSQL) MySQL Cluster is available in two Editions: Community Edition (Open Source, freely downloadable from mysql.com) Carrier Grade Edition (Commercial Edition, can be downloaded from eDelivery for evaluation purposes) MySQL Carrier Grade Edition adds on the top of the Community Edition: Commercial Extensions (MySQL Cluster Manager, MySQL Enterprise Monitor, MySQL Cluster Installer) Oracle's Premium Support Services (largest team of MySQL experts backed by MySQL developers, forward compatible hot fixes, multi-language support, and more) We concluded talking about the MySQL Cluster vision: MySQL Cluster is the default database for anyone deploying rapidly evolving, realtime transactional services at web-scale, where downtime is simply not an option. From a practical point of view the HOL's steps were: MySQL Cluster installation start & monitoring of the MySQL Cluster processes client connection to the Management Server and to an SQL Node connection using the NoSQL NDB API and the Connector J In the hope that this blog post can help you get started with MySQL Cluster, I take the opportunity to thank you for the questions you made both during the HOL and at the MySQL Cluster booth. Slides are also on SlideShares: Santo Leto - MySQL Connect 2012 - Getting Started with Mysql Cluster Happy Clustering!

    Read the article

  • What Error Messages Reveal

    - by ultan o'broin
    I love this blog entry Usability doesn't mean UI Especially the part: Ask for a list of all error messages when you do your next vendor evaluation. You will learn more about the vendor's commitment to usability and product quality than you will fathom from a slick demo. Not so sure about the part about error messages not being "hip" or "glamorous" though. I know... I should get out more...:)

    Read the article

  • What I like about WIF&rsquo;s Claims-based Authorization

    - by Your DisplayName here!
    In “traditional” .NET with its IPrincipal interface and IsInRole method, developers were encouraged to write code like this: public void AddCustomer(Customer customer) {     if (Thread.CurrentPrincipal.IsInRole("Sales"))     {         // add customer     } } In code reviews I’ve seen tons of code like this. What I don’t like about this is, that two concerns in your application get tightly coupled: business and security logic. But what happens when the security requirements change – and they will (e.g. members of the sales role and some other people from different roles need to create customers)? Well – since your security logic is sprinkled across your project you need to change the security checks in all relevant places (and make sure you don’t forget one) and you need to re-test, re-stage and re-deploy the complete app. This is clearly not what we want. WIF’s claims-based authorization encourages developers to separate business code and authorization policy evaluation. This is a good thing. So the same security check with WIF’s out-of-the box APIs would look like this: public void AddCustomer(Customer customer) {     try     {         ClaimsPrincipalPermission.CheckAccess("Customer", "Add");           // add customer     }     catch (SecurityException ex)     {         // access denied     } } You notice the fundamental difference? The security check only describes what the code is doing (represented by a resource/action pair) – and does not state who is allowed to invoke the code. As I mentioned earlier – the who is most probably changing over time – the what most probably not. The call to ClaimsPrincipalPermission hands off to another class called the ClaimsAuthorizationManager. This class handles the evaluation of your security policy and is ideally in a separate assembly to allow updating the security logic independently from the application logic (and vice versa). The claims authorization manager features a method called CheckAccess that retrieves three values (wrapped inside an AuthorizationContext instance) – action (“add”), resource (“customer”) and the principal (including its claims) in question. CheckAccess then evaluates those three values and returns true/false. I really like the separation of concerns part here. Unfortunately there is not much support from Microsoft beyond that point. And without further tooling and abstractions the CheckAccess method quickly becomes *very* complex. But still I think that is the way to go. In the next post I will tell you what I don’t like about it (and how to fix it).

    Read the article

  • Of C# Iterators and Performance

    - by James Michael Hare
    Some of you reading this will be wondering, "what is an iterator" and think I'm locked in the world of C++.  Nope, I'm talking C# iterators.  No, not enumerators, iterators.   So, for those of you who do not know what iterators are in C#, I will explain it in summary, and for those of you who know what iterators are but are curious of the performance impacts, I will explore that as well.   Iterators have been around for a bit now, and there are still a bunch of people who don't know what they are or what they do.  I don't know how many times at work I've had a code review on my code and have someone ask me, "what's that yield word do?"   Basically, this post came to me as I was writing some extension methods to extend IEnumerable<T> -- I'll post some of the fun ones in a later post.  Since I was filtering the resulting list down, I was using the standard C# iterator concept; but that got me wondering: what are the performance implications of using an iterator versus returning a new enumeration?   So, to begin, let's look at a couple of methods.  This is a new (albeit contrived) method called Every(...).  The goal of this method is to access and enumeration and return every nth item in the enumeration (including the first).  So Every(2) would return items 0, 2, 4, 6, etc.   Now, if you wanted to write this in the traditional way, you may come up with something like this:       public static IEnumerable<T> Every<T>(this IEnumerable<T> list, int interval)     {         List<T> newList = new List<T>();         int count = 0;           foreach (var i in list)         {             if ((count++ % interval) == 0)             {                 newList.Add(i);             }         }           return newList;     }     So basically this method takes any IEnumerable<T> and returns a new IEnumerable<T> that contains every nth item.  Pretty straight forward.   The problem?  Well, Every<T>(...) will construct a list containing every nth item whether or not you care.  What happens if you were searching this result for a certain item and find that item after five tries?  You would have generated the rest of the list for nothing.   Enter iterators.  This C# construct uses the yield keyword to effectively defer evaluation of the next item until it is asked for.  This can be very handy if the evaluation itself is expensive or if there's a fair chance you'll never want to fully evaluate a list.   We see this all the time in Linq, where many expressions are chained together to do complex processing on a list.  This would be very expensive if each of these expressions evaluated their entire possible result set on call.    Let's look at the same example function, this time using an iterator:       public static IEnumerable<T> Every<T>(this IEnumerable<T> list, int interval)     {         int count = 0;         foreach (var i in list)         {             if ((count++ % interval) == 0)             {                 yield return i;             }         }     }   Notice it does not create a new return value explicitly, the only evidence of a return is the "yield return" statement.  What this means is that when an item is requested from the enumeration, it will enter this method and evaluate until it either hits a yield return (in which case that item is returned) or until it exits the method or hits a yield break (in which case the iteration ends.   Behind the scenes, this is all done with a class that the CLR creates behind the scenes that keeps track of the state of the iteration, so that every time the next item is asked for, it finds that item and then updates the current position so it knows where to start at next time.   It doesn't seem like a big deal, does it?  But keep in mind the key point here: it only returns items as they are requested. Thus if there's a good chance you will only process a portion of the return list and/or if the evaluation of each item is expensive, an iterator may be of benefit.   This is especially true if you intend your methods to be chainable similar to the way Linq methods can be chained.    For example, perhaps you have a List<int> and you want to take every tenth one until you find one greater than 10.  We could write that as:       List<int> someList = new List<int>();         // fill list here         someList.Every(10).TakeWhile(i => i <= 10);     Now is the difference more apparent?  If we use the first form of Every that makes a copy of the list.  It's going to copy the entire list whether we will need those items or not, that can be costly!    With the iterator version, however, it will only take items from the list until it finds one that is > 10, at which point no further items in the list are evaluated.   So, sounds neat eh?  But what's the cost is what you're probably wondering.  So I ran some tests using the two forms of Every above on lists varying from 5 to 500,000 integers and tried various things.    Now, iteration isn't free.  If you are more likely than not to iterate the entire collection every time, iterator has some very slight overhead:   Copy vs Iterator on 100% of Collection (10,000 iterations) Collection Size Num Iterated Type Total ms 5 5 Copy 5 5 5 Iterator 5 50 50 Copy 28 50 50 Iterator 27 500 500 Copy 227 500 500 Iterator 247 5000 5000 Copy 2266 5000 5000 Iterator 2444 50,000 50,000 Copy 24,443 50,000 50,000 Iterator 24,719 500,000 500,000 Copy 250,024 500,000 500,000 Iterator 251,521   Notice that when iterating over the entire produced list, the times for the iterator are a little better for smaller lists, then getting just a slight bit worse for larger lists.  In reality, given the number of items and iterations, the result is near negligible, but just to show that iterators come at a price.  However, it should also be noted that the form of Every that returns a copy will have a left-over collection to garbage collect.   However, if we only partially evaluate less and less through the list, the savings start to show and make it well worth the overhead.  Let's look at what happens if you stop looking after 80% of the list:   Copy vs Iterator on 80% of Collection (10,000 iterations) Collection Size Num Iterated Type Total ms 5 4 Copy 5 5 4 Iterator 5 50 40 Copy 27 50 40 Iterator 23 500 400 Copy 215 500 400 Iterator 200 5000 4000 Copy 2099 5000 4000 Iterator 1962 50,000 40,000 Copy 22,385 50,000 40,000 Iterator 19,599 500,000 400,000 Copy 236,427 500,000 400,000 Iterator 196,010       Notice that the iterator form is now operating quite a bit faster.  But the savings really add up if you stop on average at 50% (which most searches would typically do):     Copy vs Iterator on 50% of Collection (10,000 iterations) Collection Size Num Iterated Type Total ms 5 2 Copy 5 5 2 Iterator 4 50 25 Copy 25 50 25 Iterator 16 500 250 Copy 188 500 250 Iterator 126 5000 2500 Copy 1854 5000 2500 Iterator 1226 50,000 25,000 Copy 19,839 50,000 25,000 Iterator 12,233 500,000 250,000 Copy 208,667 500,000 250,000 Iterator 122,336   Now we see that if we only expect to go on average 50% into the results, we tend to shave off around 40% of the time.  And this is only for one level deep.  If we are using this in a chain of query expressions it only adds to the savings.   So my recommendation?  If you have a resonable expectation that someone may only want to partially consume your enumerable result, I would always tend to favor an iterator.  The cost if they iterate the whole thing does not add much at all -- and if they consume only partially, you reap some really good performance gains.   Next time I'll discuss some of my favorite extensions I've created to make development life a little easier and maintainability a little better.

    Read the article

  • Try a sample: Using the counter predicate for event sampling

    - by extended_events
    Extended Events offers a rich filtering mechanism, called predicates, that allows you to reduce the number of events you collect by specifying criteria that will be applied during event collection. (You can find more information about predicates in Using SQL Server 2008 Extended Events (by Jonathan Kehayias)) By evaluating predicates early in the event firing sequence we can reduce the performance impact of collecting events by stopping event collection when the criteria are not met. You can specify predicates on both event fields and on a special object called a predicate source. Predicate sources are similar to action in that they typically are related to some type of global information available from the server. You will find that many of the actions available in Extended Events have equivalent predicate sources, but actions and predicates sources are not the same thing. Applying predicates, whether on a field or predicate source, is very similar to what you are used to in T-SQL in terms of how they work; you pick some field/source and compare it to a value, for example, session_id = 52. There is one predicate source that merits special attention though, not just for its special use, but for how the order of predicate evaluation impacts the behavior you see. I’m referring to the counter predicate source. The counter predicate source gives you a way to sample a subset of events that otherwise meet the criteria of the predicate; for example you could collect every other event, or only every tenth event. Simple CountingThe counter predicate source works by creating an in memory counter that increments every time the predicate statement is evaluated. Here is a simple example with my favorite event, sql_statement_completed, that only collects the second statement that is run. (OK, that’s not much of a sample, but this is for demonstration purposes. Here is the session definition: CREATE EVENT SESSION counter_test ON SERVERADD EVENT sqlserver.sql_statement_completed    (ACTION (sqlserver.sql_text)    WHERE package0.counter = 2)ADD TARGET package0.ring_bufferWITH (MAX_DISPATCH_LATENCY = 1 SECONDS) You can find general information about the session DDL syntax in BOL and from Pedro’s post Introduction to Extended Events. The important part here is the WHERE statement that defines that I only what the event where package0.count = 2; in other words, only the second instance of the event. Notice that I need to provide the package name along with the predicate source. You don’t need to provide the package name if you’re using event fields, only for predicate sources. Let’s say I run the following test queries: -- Run three statements to test the sessionSELECT 'This is the first statement'GOSELECT 'This is the second statement'GOSELECT 'This is the third statement';GO Once you return the event data from the ring buffer and parse the XML (see my earlier post on reading event data) you should see something like this: event_name sql_text sql_statement_completed SELECT ‘This is the second statement’ You can see that only the second statement from the test was actually collected. (Feel free to try this yourself. Check out what happens if you remove the WHERE statement from your session. Go ahead, I’ll wait.) Percentage Sampling OK, so that wasn’t particularly interesting, but you can probably see that this could be interesting, for example, lets say I need a 25% sample of the statements executed on my server for some type of QA analysis, that might be more interesting than just the second statement. All comparisons of predicates are handled using an object called a predicate comparator; the simple comparisons such as equals, greater than, etc. are mapped to the common mathematical symbols you know and love (eg. = and >), but to do the less common comparisons you will need to use the predicate comparators directly. You would probably look to the MOD operation to do this type sampling; we would too, but we don’t call it MOD, we call it divides_by_uint64. This comparator evaluates whether one number is divisible by another with no remainder. The general syntax for using a predicate comparator is pred_comp(field, value), field is always first and value is always second. So lets take a look at how the session changes to answer our new question of 25% sampling: CREATE EVENT SESSION counter_test_25 ON SERVERADD EVENT sqlserver.sql_statement_completed    (ACTION (sqlserver.sql_text)    WHERE package0.divides_by_uint64(package0.counter,4))ADD TARGET package0.ring_bufferWITH (MAX_DISPATCH_LATENCY = 1 SECONDS)GO Here I’ve replaced the simple equivalency check with the divides_by_uint64 comparator to check if the counter is evenly divisible by 4, which gives us back every fourth record. I’ll leave it as an exercise for the reader to test this session. Why order matters I indicated at the start of this post that order matters when it comes to the counter predicate – it does. Like most other predicate systems, Extended Events evaluates the predicate statement from left to right; as soon as the predicate statement is proven false we abandon evaluation of the remainder of the statement. The counter predicate source is only incremented when it is evaluated so whether or not the counter is incremented will depend on where it is in the predicate statement and whether a previous criteria made the predicate false or not. Here is a generic example: Pred1: (WHERE statement_1 AND package0.counter = 2)Pred2: (WHERE package0.counter = 2 AND statement_1) Let’s say I cause a number of events as follows and examine what happens to the counter predicate source. Iteration Statement Pred1 Counter Pred2 Counter A Not statement_1 0 1 B statement_1 1 2 C Not statement_1 1 3 D statement_1 2 4 As you can see, in the case of Pred1, statement_1 is evaluated first, when it fails (A & C) predicate evaluation is stopped and the counter is not incremented. With Pred2 the counter is evaluated first, so it is incremented on every iteration of the event and the remaining parts of the predicate are then evaluated. In this example, Pred1 would return an event for D while Pred2 would return an event for B. But wait, there is an interesting side-effect here; consider Pred2 if I had run my statements in the following order: Not statement_1 Not statement_1 statement_1 statement_1 In this case I would never get an event back from the system because the point at which counter=2, the rest of the predicate evaluates as false so the event is not returned. If you’re using the counter target for sampling and you’re not getting the expected events, or any events, check the order of the predicate criteria. As a general rule I’d suggest that the counter criteria should be the last element of your predicate statement since that will assure that your sampling rate will apply to the set of event records defined by the rest of your predicate. Aside: I’m interested in hearing about uses for putting the counter predicate criteria earlier in the predicate statement. If you have one, post it in a comment to share with the class. - Mike Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Débat Scrum : Que pensez-vous des certifications et processus de certifications définis par la Scrum

    Bonjour, La majorité des gens a connaissance des 2 formations certifiantes historiques autour de Scrum :Certified Scrum Product Owner Certified ScrumMaster (à laquelle s'est rajouté depuis le 1er octobre 2009 une évaluation en ligne) Ces certifications sont catégorisées Foundation-Level Certification par la Scrum Alliance. Ces derniers temps, sont apparues d'autres certifications, à savoir :Certified Scrum Developer (Mid-Level Certifications) Certifie...

    Read the article

  • Mind the gap, the latest version number for SQL Server 2008 R2 is....

    - by ssqa.net
    Since the news about SQL Server 2008 R2 RTM is publicised I have downloaded and installed the Evaluation edition and R2 Express edition. You can also download SQL Server 2008 R2 RTM - Management Studio Express (with pre-registration) The Microsoft® SQL Server® 2008 R2 RTM - Express is a powerful and reliable data management system that delivers a rich set of features, data protection, and performance for embedded applications, lightweight Web applications, and local data stores. Designed for easy...(read more)

    Read the article

  • Learning Resources for SharePoint

    - by Enrique Lima
    SharePoint 2010 Reference: Software Development Kit SharePoint 2010: Getting Started with Development on SharePoint 2010 Hands-on Labs in C# and Visual Basic SharePoint Developer Training Kit Professional Development Evaluation Guide and Walkthrough SharePoint Server 2010: Advanced Developer Training Presentations

    Read the article

  • Implement Negascout Algorithm with stack

    - by Dan
    I'm not familiar with how these stack exchange accounts work so if this is double posting I apologize. I asked the same thing on stackoverflow. I have added an AI routine to a game I am working on using the Negascout algorithm. It works great, but when I set a higher maximum depth it can take a few seconds to complete. The problem is it blocks the main thread, and the framework I am using does not have a way to deal with multi-threading properly across platforms. So I am trying to change this routine from recursively calling itself, to just managing a stack (vector) so that I can progress through the routine at a controlled pace and not lock up the application while the AI is thinking. I am getting hung up on the second recursive call in the loop. It relies on a returned value from the first call, so I don't know how to add these to a stack. My Working c++ Recursive Code: MoveScore abNegascout(vector<vector<char> > &board, int ply, int alpha, int beta, char piece) { if (ply==mMaxPly) { return MoveScore(evaluation.evaluateBoard(board, piece, oppPiece)); } int currentScore; int bestScore = -INFINITY; MoveCoord bestMove; int adaptiveBeta = beta; vector<MoveCoord> moveList = evaluation.genPriorityMoves(board, piece, findValidMove(board, piece, false)); if (moveList.empty()) { return MoveScore(bestScore); } bestMove = moveList[0]; for(int i=0;i<moveList.size();i++) { MoveCoord move = moveList[i]; vector<vector<char> > newBoard; newBoard.insert( newBoard.end(), board.begin(), board.end() ); effectMove(newBoard, piece, move.getRow(), move.getCol()); // First Call ****** MoveScore current = abNegascout(newBoard, ply+1, -adaptiveBeta, -max(alpha,bestScore), oppPiece); currentScore = - current.getScore(); if (currentScore>bestScore){ if (adaptiveBeta == beta || ply>=(mMaxPly-2)){ bestScore = currentScore; bestMove = move; }else { // Second Call ****** current = abNegascout(newBoard, ply+1, -beta, -currentScore, oppPiece); bestScore = - current.getScore(); bestMove = move; } if(bestScore>=beta){ return MoveScore(bestMove,bestScore); } adaptiveBeta = max(alpha, bestScore) + 1; } } return MoveScore(bestMove,bestScore); } If someone can please help by explaining how to get this to work with a simple stack. Example code would be much appreciated. While c++ would be perfect, any language that demonstrates how would be great. Thank You.

    Read the article

  • Effective Business Continuity Planning

    - by Chandra Vennapoosa
    While no one can be sure of where or when a disaster will occur, or what form the disaster will come in, it is important to be prepared for the unexpected. There are many companies today that have not taken into consideration the impact of disasters and this is a grave mistake. BCP Guidelines BCP for Effective Planning Building an Efficient Recovery Solution Plan Recovery Point Objective Hardware and Data Back Up Requirements Evaluation Read here :  Effective Business Continuity Planning

    Read the article

  • Oracle acquires Pillar Data Systems

    - by nospam(at)example.com (Joerg Moellenkamp)
    So far it was an investment of Larry Ellison, but now it's part of Oracle: Oracle has acquired Pillar Data Systems.. You will find more information in the press release.. As i already smell some of the comments:Pillar Data Systems is majority owned by Oracle CEO Larry Ellison. The evaluation and negotiation of the transaction was led by an independent committee of Oracle's Board of Directors. The transaction is structured as a 100% earn-out with no up-front payment.

    Read the article

  • How is precedence determined in C pointers?

    - by ankur.trapasiya
    I've come across two pointer declarations that I'm having trouble understanding. My understanding of precedence rules goes something like this: Operator Precedence Associativity (), [ ] 1 Left to Right *, identifier 2 Right to Left Data type 3 But even given this, I can't seem to figure out how to evaluate the following examples correctly: First example float * (* (*ptr)(int))(double **,char c) My evaluation: *(ptr) (int) *(*ptr)(int) *(*(*ptr)(int)) Then, double ** char c Second example unsigned **( * (*ptr) [5] ) (char const *,int *) *(ptr) [5] *(*ptr)[5] *(*(*ptr)[5]) **(*(*ptr)[5]) How should I read them?

    Read the article

< Previous Page | 6 7 8 9 10 11 12 13 14 15 16 17  | Next Page >