| Author | Message | ||
     Abilash |
Hey all, Please let me know which is the best estimate for Calculating Pla}nning or Cost of a project ?? Is It SLIM or FP ?? Also do share information on the SLIM Estimate and what is its limitation etc !! Cheers Aby | ||
| Paul Goodman |
Aby, SLIM is an estimating model for cost (pdays) and duration. FPA is a sizing technique. So the two techniques or models are complimentary rather than competitive. One use of FPA is for the size in FPts to feed estimating models be they SLIM, COCOMO, in house or others. Hope this helps, Paul | ||
| Abilash |
Thanks Paul !! Do you have any information on the SLIM Estimate ?? one of our clients use SLIM and we use FP extensively !! I just wanted to have more info on the SLIM tool !! thanks Aby | ||
| Paul Goodman |
SLIM was originally developed by Larry Putnam based on the Rayleigh distribution of effort, something often associated with R&D type projects. There is a SLIM toolset marketed by QSM in the US and Europe. European contact is Hans Vonk. I have not used SLIM recently so cannot really comment further. Hope this helps, Paul | ||
| Ashok Kumar |
What Paul says is right. However, one can use FP productivity rates for estimating project cost. I have done some study and compared FP Productivity estimates with COCOMO-II and SLiM 6.5. Please see the attached graphics. This Study was done for a main frame project for fixed configuration of SLiM and COCOMO Tools. Ashok  | ||
| Abilash |
Thanks Paul and Ashok !! Ashok ; Your Analysis says that FP is better compared to SLIM or COCOMO !! can you provide some documentation on SLIM ? It would be of great help for my analysis !! Also for calculating Size we use COCOMO or FP Count by IFPUG !! what do we really use SLIM For apart from calculating man days and man months ! cheers Aby | ||
| Ashok Kumar |
Abilash, You can get evaluation copy of SLiM tool from QSM web site. COCOMO Tool is in public domain and you can download it. Use Google to get to these sites and other related material. Regarding accuracy of estimates, it's difficult to say which one is more accurate. As my experience of estimating project cost, SLiM tend to estimate less than the COCOMO for large size projects. For medium sized projects, all three estimates are similar. For getting more accurate estimates from SLiM or COCOMO, one has to reconfigure these tools each time to take account of all special conditions of a project. Normally, I use FP productivity method for quick or ballpark estimates. Then I refine these estimates by configuring COCOMO and SLiM. My Final estimates takes inputs from other considerations that could not be covered by these tools. SLiM and COCOMO can give project phase-level and sometimes tasks group level estimates. SLiM estimates are Risk relative i.e. not absolute. FP and COCOMO based estimates are not Risk relative i.e. these are produced for 50% assurance level. SLiM can produce estimates up to 95+% assurance level. Please note here: estimates rapidly increase as assurance level is increased from 50% towards 90%. SLiM tool is very good in graphical presentation of various estimates. COCOMO can export data to Excel where you can create your own graphs. 'hope this will help you. Cheers Ashok | ||
| Paul Goodman |
Aby, My apologies in advance for being pedantic but I think we have to be careful with terminology if we are not going to confuse people. SLIM and COCOMO (and I assume the FP productivity data used by Ashok) are estimating tools or models. As they are based on different source data sets they are likely to give different estimates of effort for the same size, exactly as Ashok shows. But one is not "better" than the other unless it can be shown to be more accurate in terms of predicting the reality. In other words, a good estimating model is one where the prediction of effort (and duration perhaps) for a certain size is close to the actual effort taken. There is also the question of calibration but let us not concern ourselves with that today. We also have to be careful when we talk about sizing. COCOMO does not size. It accepts the size parameter as input either as IFPUG FPts or KLOC (I recall that FPt data in COCOMO is backfired to KLOC anyway). And of course you are correct, SLIM estimates man days or man months and elapsed time, just like COCOMO but using a different formula. If you use this link you will find lots of info on the SLIM toolset: http://www.qsm.com/products.html Again, hope this helps, Paul | ||
| Mauricio Aguiar |
Paul is right. I would like to add that one should be careful not to confuse estimating approaches with software tools. As for SLIM and COCOMO, both methods require more than a casual look to be thoroughly understood. I recommend reading *Software Cost Estimation with COCOMO II* and *Measures for Excellence - Reliable Software on Time, Within Budget* before trying to compare SLIM and COCOMO to other methods. Note that local calibration is essential when using models like COCOMO and SLIM. You may want to read the article http://sunset.usc.edu/events/2005/COCOMO/presentations/CIILocalCalibrationPaper. pdf on calibrating COCOMO II using IFPUG Function Points as input. Hope this information is helpful. Regards, Mauricio. | ||
| Abilash |
Thank You All !! Information was really helpful!! cheerz Aby |
