RE: [ai-geostats] Treatment of gold outliers from belt samples
Hello Digby, By sampling problem, I meant aspects of the collection + preparation of the eight samples from the one ton sample. From Kevin's description this all occurs at the laboratory. Thank you Isobel for your clear breakdown of the issues. The geologist in me was trying to draw attention to consideration of the sampling issues as well as the statistical. Sorry to worry you with my statement, Isobel. I was thinking that this would be the case if the bulk sample was truly homoginised. But as you rightly point out it is unlikey to be so in most crushed gold deposit samples. David Reid -Original Message- From: Digby Millikan [mailto:[EMAIL PROTECTED] Sent: Friday, 13 May 2005 7:34 PM To: ai-geostats; Reid, David W Subject: Re: [ai-geostats] Treatment of gold outliers from belt samples David, A few questions, How would you determine the number of samples? When you say sampling problem, do you mean problems in the laboratory? btw Michel David's example of the use of sichel t-estimator is the use of five samples from Harmony Mine. He also goes onto discuss pro's and con's of having 5 samples i.e. arithmetic mean of 5 samples 1195 in-dwt (47.2 gram.metres) sichel mean of 5 samples 894 in-dwt (35.3 gram metres) mean derived from 5,170 samples 618 in-dwt (24.4 gram metres) Regards Digby * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Kevin, You may also be interested in this website, if you select the applet and a skewed distribution, you can see examples of sample distributions. http://www.intuitor.com/statistics/CentralLim.html Digby * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
David, A few questions, How would you determine the number of samples? When you say sampling problem, do you mean problems in the laboratory? btw Michel David's example of the use of sichel t-estimator is the use of five samples from Harmony Mine. He also goes onto discuss pro's and con's of having 5 samples i.e. arithmetic mean of 5 samples 1195 in-dwt (47.2 gram.metres) sichel mean of 5 samples 894 in-dwt (35.3 gram metres) mean derived from 5,170 samples 618 in-dwt (24.4 gram metres) Regards Digby * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Richard, If you wish to use the chip samples to estimate the mean the formulea for the estimate of the mean of a lognormal population is; T* = exp(u* + 0.5 s*^2) pp73 Practical Geostatistics 2000 I. Clark i.e. you need to have the variance of the samples to estimate the mean. The problem is that the variance of the chip samples is different from the variance of the blocks they represent, so before plugging the s* into the equation, you have to adjust it for it's support. You can do this using the volume-variance charts provided in Mining Geostatistics, A.Journel. Make sense? Regards Digby * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
RE: [ai-geostats] Treatment of gold outliers from belt samples
Hi Digby, Just a hunch that it is sampling problem. When dealing with course gold particles a representative sample size is rarely practical. Without knowledge of gold particle size and sub-sample/analyte mass it is hard to make this judgement. As you point out, the sub-sample values should have a normal distribution. Increasing the number of samples (n) would help. Eight samples is unlikely to be enough to characterise what is probably a highly variable distribution. Regards David Reid * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Digby, What do you mean "You also may wish to consider volume-variance relationships when comparing the chip sample means and the conveyor belt means, as the sample sizes are different"? The volume-variance relationship does not affect the mean - it is concerned with the changes in variance at different support (volume) sizes. If you sample the belt using a pair of tweezers and large shovels then the means of each of these group of samples should be the same (given sampling is carried out correctly) - it will be the variance that changes (higher variance in the smaller (tweezer) samples, lower variance in the larger (shovel) samples). Cheers Richard Hague - Original Message - >Digby Millikan <[EMAIL PROTECTED]> >To: ai-geostats , Kevin Lowe (Office Park)><[EMAIL PROTECTED]> >Subject: Re: [ai-geostats] Treatment of gold outliers from belt samples >Date: Fri, 13 May 2005 15:23:12 +0930 >Attachments: message-footer.txt,Size: 292 bytes. >Click here to clean up the attachments on mail2webServer >Treatment of gold outliers from belt samples It seems your ore on the belt >may have mixed up a bit since being part of the >orebody, and also only represents a small part of the orebody. According to >the central limit thereom "if a series of samples, of size n, are taken >from a population (not necessarily normal) with mean u, and standard >deviation s, then the sample means will form a distribution which tends to >the normal distribution as n increases whatever the population >distribution." I have a copy of Sichel's paper and a sample >of the mean estimate calculation, which I can send you. You also may wish >to consider volume-variance relationships when comparing the chip sample >means and the conveyor belt means, as the sample sizes are different. > >Regards Digby mail2web - Check your email from the web at http://mail2web.com/ . * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Treatment of gold outliers from belt samplesDavid, Why do you think this is a sampling problem? Regards Digby - Original Message - From: Reid, David W To: ai-geostats@unil.ch Sent: Friday, May 13, 2005 11:01 AM Subject: RE: [ai-geostats] Treatment of gold outliers from belt samples Hi Kevin, The arithmetic mean is probably not a very good measure of central tendancy with this skewed data. An alternative such as that suggested by Isobel Clark may be more realistic. However it may also be benifical to investigate the root cause. This would seem to be a sampling problem. Is it possible to increase the size of the the eight samples (Preferably both total split for pulverising + analyte weight) OR to reduce the particle size of the 1 ton sample prior to splitting? Because gold is maliable it often difficult to reduce the partical size, often you only succeed in flattening the gold particles! Regards David Reid -- * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Treatment of gold outliers from belt samples It seems your ore on the belt may have mixed up a bit since being part of the orebody, and also only represents a small part of the orebody. According to the central limit thereom "if a series of samples, of size n, are taken from a population (not necessarily normal) with mean u, and standard deviation s, then the sample means will form a distribution which tends to the normal distribution as n increases whatever the population distribution." I have a copy of Sichel's paper and a sample of the mean estimate calculation, which I can send you. You also may wish to consider volume-variance relationships when comparing the chip sample means and the conveyor belt means, as the sample sizes are different. Regards Digby * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
RE: [ai-geostats] Treatment of gold outliers from belt samples
Title: Treatment of gold outliers from belt samples Hi Kevin, The arithmetic mean is probably not a very good measure of central tendancy with this skewed data. An alternative such as that suggested by Isobel Clark may be more realistic. However it may also be benifical to investigate the root cause. This would seem to be a sampling problem. Is it possible to increase the size of the the eight samples (Preferably both total split for pulverising + analyte weight) OR to reduce the particle size of the 1 ton sample prior to splitting? Because gold is maliable it often difficult to reduce the partical size, often you only succeed in flattening the gold particles! Regards David Reid * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
Re: [ai-geostats] Treatment of gold outliers from belt samples
Hi Kevin Can I refer you to the works of Herbert Sichel which was developed exactly for this problem, earliest paper Trans Inst Min Metall 1949. Or you can download my 1987 SAIMM paper from http://uk.geocities.com/drisobelclark/resume which describes Sichel's work. Isobel http://geoecosse.bizland.com"Kevin Lowe (Office Park)" <[EMAIL PROTECTED]> wrote: Hi, How should one treat obvious gold grade outliers from samples collected from a belt? The sampling is carried out by an automatic belt sampler prior to the ore being milled. The samples are collected and stored in a bin until there is approximately 1 ton of sample. The bin is then sent off to a lab which crushes and splits the 1 ton bin sample to produce 8 separate samples which are then assayed. Assuming there are no issues with the lab procedures, how should one treat a very high value? For example purposes, say the 8 samples returned grades (g/t) of 2.8, 4.6, 5.2, 4.5, 35.6, 3.6, 4.2, 4.7. The arithmetic mean for the eight is 8.15g/t but if the one high grade is removed the arithmetic mean is 4.23g/t. Should I simply exclude the high value or should I cut the value of the sample to some arbitrary value (say the upper 95% confidence limit)? Although individual chip samples collected from the orebody, for the purposes of evaluation, are highly skewed, the samples from the bin approximate a normal distribution (excluding the high value). I look forward to any comments or perhaps direction to papers or web sites on this topic. Many Thanks Kevin Lowe This e-mail message has been scanned for Viruses and Content and cleared by NetIQ MailMarshal * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm )* To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED]Signoff ai-geostats* By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
RE: [ai-geostats] Treatment of gold outliers from belt samples
Title: Treatment of gold outliers from belt samples Hi Kevin, In the case of block estimation, I do not tend to remove or cut-back outliers to a known value – these high grade samples are after all a unique feature of the deposit, they must mean something right? I do, however, tend to limit their influence on blocks being estimated. Typically, I do this by ignoring any sample above a threshold lying outside a nominated search ellipsoid. I am sure there are many other ways of treating these samples, one is the 95% limit you mention, so it will be interesting to see the various responses. Regards, Colin From: Kevin Lowe (Office Park) [mailto:[EMAIL PROTECTED] Sent: 12 May 2005 11:41 To: ai-geostats@unil.ch Subject: [ai-geostats] Treatment of gold outliers from belt samples Hi, How should one treat obvious gold grade outliers from samples collected from a belt? The sampling is carried out by an automatic belt sampler prior to the ore being milled. The samples are collected and stored in a bin until there is approximately 1 ton of sample. The bin is then sent off to a lab which crushes and splits the 1 ton bin sample to produce 8 separate samples which are then assayed. Assuming there are no issues with the lab procedures, how should one treat a very high value? For example purposes, say the 8 samples returned grades (g/t) of 2.8, 4.6, 5.2, 4.5, 35.6, 3.6, 4.2, 4.7. The arithmetic mean for the eight is 8.15g/t but if the one high grade is removed the arithmetic mean is 4.23g/t. Should I simply exclude the high value or should I cut the value of the sample to some arbitrary value (say the upper 95% confidence limit)? Although individual chip samples collected from the orebody, for the purposes of evaluation, are highly skewed, the samples from the bin approximate a normal distribution (excluding the high value). I look forward to any comments or perhaps direction to papers or web sites on this topic. Many Thanks Kevin Lowe This e-mail message has been scanned for Viruses and Content and cleared by NetIQ MailMarshal * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
[ai-geostats] Treatment of gold outliers from belt samples
Title: Treatment of gold outliers from belt samples Hi, How should one treat obvious gold grade outliers from samples collected from a belt? The sampling is carried out by an automatic belt sampler prior to the ore being milled. The samples are collected and stored in a bin until there is approximately 1 ton of sample. The bin is then sent off to a lab which crushes and splits the 1 ton bin sample to produce 8 separate samples which are then assayed. Assuming there are no issues with the lab procedures, how should one treat a very high value? For example purposes, say the 8 samples returned grades (g/t) of 2.8, 4.6, 5.2, 4.5, 35.6, 3.6, 4.2, 4.7. The arithmetic mean for the eight is 8.15g/t but if the one high grade is removed the arithmetic mean is 4.23g/t. Should I simply exclude the high value or should I cut the value of the sample to some arbitrary value (say the upper 95% confidence limit)? Although individual chip samples collected from the orebody, for the purposes of evaluation, are highly skewed, the samples from the bin approximate a normal distribution (excluding the high value). I look forward to any comments or perhaps direction to papers or web sites on this topic. Many Thanks Kevin Lowe This e-mail message has been scanned for Viruses and Content and cleared by NetIQ MailMarshal * By using the ai-geostats mailing list you agree to follow its rules ( see http://www.ai-geostats.org/help_ai-geostats.htm ) * To unsubscribe to ai-geostats, send the following in the subject or in the body (plain text format) of an email message to [EMAIL PROTECTED] Signoff ai-geostats
