My Blog Posts

Printer woes

Sat, 28 Nov 2015 06:55:57 +0000

The following lines reflect my attempts to get my new printer to do my bidding.

Conditions held constant

Printer name

Lexmark E350d

Connection

usb://Lexmark/E350d?serial=62350RR

Printer management software

CUPS 1.7.2

driver name	settings						output of lpoptions	tray used	result of printing a page
	share over net	color management	trays	duplex	flash	storage			from okular	from Firefox	from shell with lp
Lexmark E350d en	yes	no	1+2	yes	yes	160 MB	copies=1 device-uri=usb:Lexmark/E350d?serial=62350RR finishings=3 job-hold-until=no-hold job-priority=50 job-sheets=none,none marker-change-time=0 number-up=1 ppd-timestamp=* printer-commands=ReportLevels printer-info='Lexmark Lexmark E350d' printer-is-accepting-jobs=true printer-is-colormanaged=false printer-make-and-model='Lexmark E350d' printer-state=3 printer-state-change-time=1448606941 printer-state-reasons=none printer-type=8425684 printer-uri-supported=ipp:localhost:631/printers/Lexmark_Lexmark_E350d sides=two-sided-long-edge	Standard	works (doesn't recognise re-filled tray though)	as text	works

by derwisch

2048 Racket

Sun, 11 Oct 2015 11:24:38 +0000

by derwisch

An analysis of Patchwork tiles

Sun, 19 Jul 2015 03:54:09 +0000

Patchwork is a 2-player game by Uwe Rosenberg which uses known mechanisms (tile-fitting, a time row, a circular card row, economics with multiple resources) and combines them in a refreshing way. Like many 2-player games, it can get confrontational as denying your opponent nice things is as good as getting nice things yourself. There has been some discussion about which tiles are "nice", which ones are adequately priced, and which ones are the real bargains.

The cost of tiles in buttons should somehow be dependent on their other features, which are their size and shape, their cost in time, and their button yield. So let's look at some of the features, and see if statistics can help us there. Images are clickable to inspect them at their original size.

Return of investment, and how time is money

Button yield should be a major factor. A patch purchased on your first turn pays off eight times. If only I could afford that fancy piece of cloth adorned with three buttons. My start capital of a mere five buttons, however, only permits me to purchase half of the double button producers, and none of the triple-buttoned cash cows. Let's see how purchase price relates to expected gain.

There is quite an association, as could be expected.

One other factor determining the price of a patch should be the time that goes into sewing it into the design. You have 59 spaces to cover before your 9 by 9 quilt has to be finished, so all things being equal, the lower the time, the higher that cost should be.

Most illustrative is a pair of tiles where all things are equal, and that's the two corner triominoes, which have a time and cost of (3, 1) and (1, 3), respectively.

Let's see how the correspondence between time and cost looks in general.

Now this is a lot less clear-cut. If anything, time is positively associated with cost. How can that be? Apparently, other factors, namely the button yield, play a greater role and are masking the effect here. Let's separate the plots by their button yield.

This is a lot more insightful now. We can now apply linear regression to our data and see how exactly time and yield affect button prices. A linear regression of patch costs on button yield and time used scores an adjusted R² of 77 per cent, which means 77 per cent of the variance can be explained by those to cost factors. Here are the effect estimates with their confidence intervals.

parameter	estimate	2.5 %	97.5 %
(Intercept)	3.643	2.497	4.789
buttons	3.444	2.725	4.162
time	-1.098	-1.573	-0.624

We can see that each button sewn on a patch hikes up the price by about 3.4 buttons … which means that those patches are worth acquiring as long as your time counter has at least four buttons ahead of it. Time and button cost seem to have similar weight, as was exemplified by those two corner triominoes.

I think this is the moment to remind ourselves that most tiles are situational. Sometimes you need badly to hike up your income, lest you'd be screwed for choice for the next few moves. Sometimes you want to take *more* time to sew that patch in to gain that leather patch before your opponent, conveniently fixing a gap. Whatever we end up with in this analysis, it doesn't spare us thinking about what is most helpful in the current situation. The idea of the exercise is rather to get an idea about how Uwe got to price the tiles the way they are.

Size does matter, or does it?

While 77 per cent explained variance is pretty good (and would have me gloating and raving if I achieved this at my job working with medical data) we notice that there are cost gaps between similar tiles left. So what could be another factor? Size, of course, as each empty square eats two of your hard-earned buttons at the end of the game. So tile size should count for a lot in linear regression, right?

parameter	estimate	2.5 %	97.5 %
(Intercept)	3.6	1.609	5.59
buttons	3.443	2.711	4.175
time	-1.1	-1.587	-0.613
size	0.01	-0.356	0.376

Wrong.

Size does not matter.

But how can that be? Possibly Uwe thought that big tiles are more situational and harder to fit in? Let's think of a measure that reflects the "awkwardness" of a tile. The least awkward one is the rectangle. So why not draw a rectangular bounding box around each tile and count the number of squares outside the tile, the number of "holes" in the shape? The "H" heptomino would have two holes by that measure, the only octomino in play four, and the cactus hexomino is discounted by its six holes which make it hard to fit in, placing its cost at the price of a measly domino with no holes. So let's see how the number of holes affects the price.

parameter	estimate	2.5 %	97.5 %
(Intercept)	2.502	0.725	4.28
buttons	3.606	2.985	4.227
time	-1.22	-1.634	-0.806
size	0.515	0.1	0.93
holes	-0.467	-0.725	-0.209

And lo and behold, not only does the number of holes affect the price, but when considering the number of holes in addition to size, size enters the equation all of a sudden!

Another thing could be considered. Some tile shapes are invariant to flipping or rotating. Some of them don't change their appearance at all, no matter how you twist and turn them (I'm looking at you, x pentomino). Some of the tiles are versa-tiles, others stubbornly cling to one way to be built in. Why not differentiate the tiles according to the number of different shapes they can assume, i.e. 1, 2, 4, or 8? A higher versatility would not necessarily mean a higher use in your current situation, as you can only fit the patch in one way, but it means it is of use more often. Let's see how versatility influences pricing.

parameter	estimate	2.5 %	97.5 %
(Intercept)	2.544	0.417	4.671
buttons	3.608	2.973	4.242
time	-1.219	-1.642	-0.796
size	0.514	0.09	0.937
holes	-0.467	-0.73	-0.203
orient	-0.015	-0.405	0.376

Again, the answer is "not very much". Nor does it affect the influence of the shape parameters "size" and "holes", which are more or less invariant to the inclusion of the orientation parameter. But maybe I have forgotten another term associated with price whose inclusion makes the orientation parameter stand out.

Bargains, and button sinks

So, which are the bargains and which patches would we rather avoid? This can be answered by a technique called "residual plot", which is a scatter plot of the deviation of the price from the price we expect if we believe in the model (which is, in this case, the model without the "orient" parameter). The x axis of the plot reflects the expected price.

What we can see is that the latin cross hexomino, given away for free, is worth every button spent on it, and more. There are two tiles which the model would suggest could have been given away for free. Those are the "Z" hexomino with its eight holes and the corner triomino eating three of your precious time. The most extortionate price seems to be 10 buttons for the long L hexomino. In this case, however, the "number of holes" metric seems to be a bit harsh as the "L" shape is rarely unwelcome. Even so, more than half of the patches (within the zone between -.5 and .5) seem to be adequately priced, as raising or lowering the price by one button would only increase their contribution to the total variance.

Conclusion

There are a couple of interesting lessons I see here:

Size looks underpriced, if you consider the value of two buttons per unfinished square. The large tiles come with some downsides, however. One of them is than once you have obtained the 7 by 7 prize, you can't fit more than one tile with a 3 by 3 or larger bounding box in. It seems to me that forgoing the 7 by 7 in favour of collecting large tiles up to the end game is a viable strategy.
Even in simple models like linear regression it is possible that the effect of some parameters may mask or confound the effect of other parameters. Multiple predictors may influence each other in a non-obvious way. Medical literature [is http://www.ncbi.nlm.nih.gov/pubmed/22106764] rife with predictive models built by selecting the parameters by their influence in simple models where no other explanatory variables are considered. With that approach, we'd have been stuck here with button yield and time as meaningful explanatory variables, with time showing an effect in the marginal model which is *opposite* to the effect in all models with button yield included. A great example of ecological fallacy, by the way.

by derwisch

Maperitive woes

Sat, 22 Dec 2012 06:38:29 +0000

No discussion of make(1) would be complete without an acknowledgement that it includes one of the worst design botches in the history of Unix. The use of tab characters as a required leader for command lines associated with a production means that the interpretation of a makefile can change drastically on the bases of invisible whitespace.

— Eric S. Raymond, The Art of Unix Programming

And then a few weeks later I had a user population of about a dozen, most of them friends, and I didn't want to screw up my embedded base. The rest, sadly, is history.

— Stuart Feldman, author of make, cited in Eric S. Raymond, The Art of Unix Programming

Maperitive is a program that draws maps from geodatabases. It exports to pixel- and vector-based formats. It has an own database of topographic properties of the surface of our planet, which can be rendered in the map as hillshading or contours. It does things right by separating the data layer from the presentation layer using the concepts of features in its rendering rules.

And then, there is the syntax rules.

The following is a collection of my attempts to set up a set of rendering rules, and the error messages Maperitive responds with.

Rules file	error messages	comment
Hospital-DOS-f-h.mrules	no error	"hospital" as a child of "features", line ends with x0D, x0A
Hospital-DOS-f-n-h.mrules	no error	"hospital" as a grandchild of "features", child of "nodes", line ends with x0D, x0A
Hospital-DOS-f-n-h.mrules	Error: unknown feature type "nodes", line=6, col=2	the same file as before, parsed later the same day
Hospital-DOS-f-n-h.mrules	no error	Error: unknown feature type "nodes", line=6, col=2	added a "b" in comment

The last but one file used to work, the last one not. Output from "diff Hospital-DOS-f-n-h-2.mrules Hospital-DOS-f-h-n.mrules":

4c4 < // -- no node feature; hospital as feature below "nodes" --- > // -- no node feature; hospital as feature elow "nodes"

Maperitive version 2.3.19.

Remembering this is a program developed under a Windows platform, I stopped and restarted it. Now the error message appears for the third and last incarnation, but not for the first.

Output from "diff Hospital-DOS-f-h.mrules Hospital-DOS-f-h-n.mrules":

4c4 < // -- no node feature; hospital as single feature --- > // -- no node feature; hospital as feature below "nodes" 6c6,7 < hospital : amenity=hospital --- > nodes > hospital : amenity=hospital 26,27c27 < font-family : Cambria < font-style : italic --- > font-family : Inconsolata

These lines will be extended until I find out what is going on or my patience expires, whichever comes first.

by derwisch

Polysleuth: Rules

Sun, 09 Dec 2012 21:55:04 +0000

Introduction

Polysleuth is a variation of Sid Sackson's deduction game classic "Sleuth". Players try to infer the distribution of cards between their opponents with help of their own cards and questions. The player who scores the most sleuth points wins. Game can be played with 3 to 7 players.

Material and initial settings

There are 36 cards in play, each being a unique combination of the traits Gem (Diamond, Pearl, Opal), Colour (Red, Green, Blue, Yellow) and Number (1, 2, 3). Cards are evenly and randomly distributed between players (and revealed to the individual player), any remaining cards are not distributed (they "remain in the safe"). Each player gains 30 Private Eyes at the start of the game (the currency used to place value on an individual question). The game master may impose a restriction to the randomization such as "no player has more than two cards that share more than one trait", but has to reveal this information.

Turns

All players make their move concurrently and independent from one another. A turn consists of the following moves:

* Guessing cards
* Placing questions
* Collecting Private Eyes

Turns are not revealed to public but sent as private messages to the game master.

Guessing cards

A player may guess as many cards as she desires. A guess is placed in the form: "<guess number>: <name of player> has <Number> <Colour>; <Gem>" or "<guess number>: <Number> <Colour> <Gem> is in the safe." Every player may place a guess on any one card only once in the game.

Placing questions

A player may place as many questions as she desires. A question must be assigned a weight in number of Private Eyes spent. A question must be placed in the form "<question number> (<#Private Eyes>): <name of player> has how many <trait1> {<trait2>}?" A question must address one or two traits.

Collecting Private Eyes

At the end of their turns, each player collects 12 Private Eyes.

Resolving turns

The game master answers all guesses publicly, revealing, however, only the player and guess number. The game master assigns Sleuth Points for each guess. If a card is guessed correctly for the first time, the player collects Sleuth Points even to the number of players, each subsequent guess on the card is rewarded with one point less each. If a card is guessed correctly by several players, they collect the total score, divided by the number of guessing players. For instance, in a six player game, if a card is guessed correctly for the first time by three players, they receive (6+5+4)/3 = 5 Sleuth Points each. The guessed stone is not revealed, neither is the number of points the players scored for a specific question (but see down for the total Sleuth Point score).

The game master answers for each player one question, namely the one that was assigned the most weight. Any question for one trait only is marked up by 2 Private Eyes before weights are compared. Tie breakers are: Player with less Private Eyes wins, player with less Sleuth Points wins, a coin is flipped. The game master publicises the answers in the following form: "<name of player> has <number of cards> <Trait1> {<Trait2>} (question by <name of player>, <#Private Eyes spent plus markup>)." Additionally, if a question for two traits was answered, the game master reveals to the player who asked the question the cards in question. The player still has to "guess" these cards in the subsequent turn in order to score.

The game master also publicises the Private Eye and Sleuth Point score for each player.

If no Sleuth Points can be scored any more for a specific card, its owner is publicised too.

End of Game

The game ends once all cards are publicised. Player with most Sleuth Points wins.

by derwisch

USB problems on an Ubuntu box

Sun, 26 Feb 2012 20:03:09 +0000

Preamble

My Ubuntu box displays very few connected devices automatically. As this Wiki hoster provides some ways to markup I'll use this site to document what I have tried.

The computer runs Oneiric Ocelot.

Plugging an external drive

hardware

Seagate FreeAgent external hard drive

USB plug

upper left

hotplug reaction

recognised, three options offered

dmesg

[122511.584363] usb 2-3: new high speed USB device number 3 using ehci_hcd [122511.738460] scsi9 : usb-storage 2-3:1.0 [122512.740252] scsi 9:0:0:0: Direct-Access Seagate FreeAgent 102D PQ: 0 ANSI: 4 [122512.772605] sd 9:0:0:0: Attached scsi generic sg6 type 0 [122522.130355] sd 9:0:0:0: [sdf] 976773168 512-byte logical blocks: (500 GB/465 GiB) [122522.131710] sd 9:0:0:0: [sdf] Write Protect is off [122522.131722] sd 9:0:0:0: [sdf] Mode Sense: 1c 00 00 00 [122522.133096] sd 9:0:0:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [122522.367769] sdf: sdf1 [122522.374310] sd 9:0:0:0: [sdf] Attached SCSI disk

lsusb

... Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 004 Device 002: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse Bus 002 Device 003: ID 0bc2:3001 Seagate RSS LLC

Plugging an external drive

hardware

Seagate FreeAgent external hard drive

USB plug

lower left

hotplug reaction

recognised, three options offered

dmesg

[122793.420349] usb 8-1: new high speed USB device number 4 using xhci_hcd [122793.448743] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.450359] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.451988] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.453610] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.457739] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.459360] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [122793.459809] scsi10 : usb-storage 8-1:1.0 [122794.464266] scsi 10:0:0:0: Direct-Access Seagate FreeAgent 102D PQ: 0 ANSI: 4 [122794.501259] sd 10:0:0:0: Attached scsi generic sg6 type 0 [122799.973262] sd 10:0:0:0: [sdf] 976773168 512-byte logical blocks: (500 GB/465 GiB) [122799.974529] sd 10:0:0:0: [sdf] Write Protect is off [122799.974542] sd 10:0:0:0: [sdf] Mode Sense: 1c 00 00 00 [122799.975928] sd 10:0:0:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [122799.986341] sdf: sdf1 [122799.991787] sd 10:0:0:0: [sdf] Attached SCSI disk

lsusb

Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 004 Device 002: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse Bus 008 Device 004: ID 0bc2:3001 Seagate RSS LLC

So the bus has changed to 008.

Plugging a card reader

hardware

6in1 card reader

USB plug

upper left

hotplug reaction

none

dmesg

[123036.928361] usb 4-3: new full speed USB device number 3 using ohci_hcd [123037.102163] scsi11 : usb-storage 4-3:1.0 [123038.106796] scsi 11:0:0:0: Direct-Access Generic USB Storage-SMC 0207 PQ: 0 ANSI: 0 CCS [123038.112796] scsi 11:0:0:1: Direct-Access Generic USB Storage-CFC 0207 PQ: 0 ANSI: 0 CCS [123038.118804] scsi 11:0:0:2: Direct-Access Generic USB Storage-MMC 0207 PQ: 0 ANSI: 0 CCS [123038.124796] scsi 11:0:0:3: Direct-Access Generic USB Storage-MSC 0207 PQ: 0 ANSI: 0 CCS [123038.145335] sd 11:0:0:0: Attached scsi generic sg6 type 0 [123038.145834] sd 11:0:0:1: Attached scsi generic sg7 type 0 [123038.146250] sd 11:0:0:2: Attached scsi generic sg8 type 0 [123038.146624] sd 11:0:0:3: Attached scsi generic sg9 type 0 [123069.128462] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123100.136368] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123131.240370] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123131.473816] sd 11:0:0:1: [sdg] Attached SCSI removable disk [123131.497832] sd 11:0:0:0: [sdf] Attached SCSI removable disk [123131.547883] sd 11:0:0:3: [sdi] Attached SCSI removable disk [123162.220051] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123193.192364] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123224.168110] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123255.144375] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123286.184364] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123317.224360] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123348.200379] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123379.176394] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123410.152355] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123441.192367] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123472.232056] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123503.208367] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123534.184088] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123565.160364] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123596.136364] usb 4-3: reset full speed USB device number 3 using ohci_hcd [123596.310034] sd 11:0:0:2: [sdh] READ CAPACITY failed [123596.310046] sd 11:0:0:2: [sdh] Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK [123596.310056] sd 11:0:0:2: [sdh] Sense not available. [123596.326027] sd 11:0:0:2: [sdh] Write Protect is off [123596.326039] sd 11:0:0:2: [sdh] Mode Sense: 00 00 00 00 [123596.342014] sd 11:0:0:2: [sdh] Asking for cache data failed [123596.342027] sd 11:0:0:2: [sdh] Assuming drive cache: write through

It seems to be associated with me having an old card reader unprepared for bigger capacity cards.

After plugging a 512 MB card, dmesg reads:

[69169.497407] sd 11:0:0:2: [sdi] Unhandled sense code [69169.497418] sd 11:0:0:2: [sdi] Result: hostbyte=invalid driverbyte=DRIVER_SENSE [69169.497428] sd 11:0:0:2: [sdi] Sense Key : Medium Error [current] [69169.497438] sd 11:0:0:2: [sdi] Add. Sense: Recorded entity not found [69169.497449] sd 11:0:0:2: [sdi] CDB: Read(10): 28 00 00 0f 2c 00 00 00 01 00 [69169.497469] end_request: critical target error, dev sdi, sector 994304

multiple times, but GUI offers the three options as expected.

lsusb

... Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 004 Device 002: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse Bus 004 Device 003: ID 0aec:3050 Neodio Technologies Corp. ND3050 8-in-1 Card Reader

Plugging a GPS device

hardware

Garmin eTrex 30 (in mass storage mode)

USB plug

lower left

hotplug reaction

recognised, three options offered for each "1.8 GiB medium" and "Garmin GARMIN Flash"

dmesg

[ 3010.528374] usb 8-1: new full speed USB device number 6 using xhci_hcd [ 3010.563183] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [ 3010.566917] xhci_hcd 0000:00:10.1: WARN: short transfer on control ep [ 3010.573896] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.577175] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.579892] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.582897] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.585903] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.589176] xhci_hcd 0000:00:10.1: WARN: Stalled endpoint [ 3010.589653] scsi8 : usb-storage 8-1:1.0 [ 3011.593166] scsi 8:0:0:0: Direct-Access Garmin GARMIN Flash 1.00 PQ: 0 ANSI: 5 [ 3011.594101] scsi 8:0:0:1: Direct-Access Garmin GARMIN Card 1.00 PQ: 0 ANSI: 5 [ 3011.704953] sd 8:0:0:0: Attached scsi generic sg6 type 0 [ 3011.705342] sd 8:0:0:1: Attached scsi generic sg7 type 0 [ 3011.716073] sd 8:0:0:0: [sdf] 3907584 512-byte logical blocks: (2.00 GB/1.86 GiB) [ 3011.717011] sd 8:0:0:1: [sdg] 3862528 512-byte logical blocks: (1.97 GB/1.84 GiB) [ 3011.718542] sd 8:0:0:0: [sdf] Write Protect is off [ 3011.718550] sd 8:0:0:0: [sdf] Mode Sense: 23 00 00 00 [ 3011.719514] sd 8:0:0:1: [sdg] Write Protect is off [ 3011.719521] sd 8:0:0:1: [sdg] Mode Sense: 23 00 00 00 [ 3011.720560] sd 8:0:0:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 3011.721516] sd 8:0:0:1: [sdg] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA [ 3011.738907] sdf: [ 3011.747116] sdg: sdg1 [ 3011.758352] sd 8:0:0:0: [sdf] Attached SCSI removable disk [ 3011.759452] sd 8:0:0:1: [sdg] Attached SCSI removable disk

lsusb

... Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 008 Device 006: ID 091e:2519 Garmin International Bus 004 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse

Plugging a memory stick

hardware

USB plug

upper left

hotplug reaction

recognised, three options offered

dmesg

[52987.200386] sd 9:0:0:0: [sdf] Synchronizing SCSI cache [52987.200878] sd 9:0:0:0: [sdf] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK [52987.224173] sd 9:0:0:1: [sdg] Synchronizing SCSI cache [52987.224214] sd 9:0:0:1: [sdg] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK [52992.524143] usb 2-3: new high speed USB device number 4 using ehci_hcd [52992.790663] scsi10 : usb-storage 2-3:1.0 [52993.791400] scsi 10:0:0:0: Direct-Access Generic Flash Disk 8.07 PQ: 0 ANSI: 2 [52993.896778] sd 10:0:0:0: Attached scsi generic sg6 type 0 [52993.898049] sd 10:0:0:0: [sdg] 15740928 512-byte logical blocks: (8.05 GB/7.50 GiB) [52993.898543] sd 10:0:0:0: [sdg] Write Protect is off [52993.898552] sd 10:0:0:0: [sdg] Mode Sense: 03 00 00 00 [52993.901217] sd 10:0:0:0: [sdg] No Caching mode page present [52993.901226] sd 10:0:0:0: [sdg] Assuming drive cache: write through [52993.905657] sd 10:0:0:0: [sdg] No Caching mode page present [52993.905668] sd 10:0:0:0: [sdg] Assuming drive cache: write through [52994.384997] sdg: sdg1 [52994.388182] sd 10:0:0:0: [sdg] No Caching mode page present [52994.388195] sd 10:0:0:0: [sdg] Assuming drive cache: write through [52994.388203] sd 10:0:0:0: [sdg] Attached SCSI removable disk

lsusb

Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 002 Device 004: ID 058f:6387 Alcor Micro Corp. Transcend JetFlash Flash Drive Bus 004 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse

Plugging a USB Hub

hardware

USB memory stick over D-Link DUB H4

USB plug

upper left

hotplug reaction

none

dmesg

[56277.672162] usb 2-3: USB disconnect, device number 4 [56283.820074] usb 2-3: new high speed USB device number 5 using ehci_hcd [56283.955441] hub 2-3:1.0: USB hub found [56283.955785] hub 2-3:1.0: 4 ports detected [56289.757130] usb 2-3.1: new high speed USB device number 6 using ehci_hcd [56289.984633] usb 2-3.1: rejected 1 configuration due to insufficient available bus power [56289.984643] usb 2-3.1: no configuration chosen from 1 choice

There is an extra power plug on the USB hub which I didn't use.

lsusb

... Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 002 Device 005: ID 05e3:0608 Genesys Logic, Inc. USB-2.0 4-Port HUB Bus 002 Device 006: ID 058f:6387 Alcor Micro Corp. Transcend JetFlash Flash Drive Bus 004 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse

Plugging a camera

hardware

Casio EX-Z350

USB plug

upper left

hotplug reaction

none

dmesg

no messages

lsusb

... Bus 001 Device 002: ID 058f:6364 Alcor Micro Corp. Bus 004 Device 003: ID 046d:c03d Logitech, Inc. M-BT96a Pilot Optical Mouse

Skeleton

hardware

USB plug

hotplug reaction

dmesg

lsusb

Conclusion

Send in camera, buy current card reader (or try card slot on desktop box), power connect USB port.

by derwisch

Food consumption in different countries

Thu, 15 Apr 2010 11:53:50 +0000

The New York Times presented a bar chart on food consumption in different countries. Here is a correspondence analysis on the distribution

, and here is the R code needed to produce it:

food <- data.frame(list(US = c(63, 298, 108, 12, 77, 47, 183, 194, 124, 88, 97), SP = c(77, 315, 149, 10, 67, 28, 112, 192, 218, 59, 152), FR = c(54, 265, 175, 20, 77, 19, 130, 184, 91, 56, 131), MX = c(45, 146, 433, 9, 14, 19, 37, 156, 203, 116, 122), JP = c(75, 147, 53, 32, 17, 19, 251, 278, 126, 94, 146), RU = c(49, 127, 144, 16, 25, 24, 70, 335, 88, 88, 125), BR = c(51, 117, 81, 12, 6, 9, 115, 208, 161, 103, 205), SA = c(28, 88, 117, 5, 14, 10, 42, 145, 64, 67, 81), CN = c(24, 31, 13, 9, 5, 5, 29, 609, 132, 65, 228), IN = c( 1, 25, 7, 1, 2, 1, 2, 207, 80, 72, 19) ), row.names = c("sauces", "dairy", "bakery", "pasta", "canned", "snacks", "conven", "veg", "fruit", "nuts", "meat")) library(MASS) cspfood <- corresp(food, nf=2) png("NYTfood.png") plot(cspfood) dev.off()

Note how the food categories marked red/orange are well separated from the ones marked blue/green in the plot. Contrary to what I've expected, France and the U.S. are quite close to one another.

by derwisch

Digitise graphs with R

Sat, 06 Mar 2010 06:14:34 +0000

Part of deciding on the size of an experiment is literature search. Has the suggested response variable been used before? (Most possibly.) What rate of an event is to be expected? What distribution does the response have? More often than not, however, especially the last question cannot be reliably answered, as many articles report effect size, p-value, period. If the hapless statistician is lucky, he will see the desired data in a box-and-whiskers plot, a dot chart or a scatter plot. A couple of times I have found myself trying to recover the original data using a ruler.

Now Thimotee has shown how to use R to digitise points from images. Something I'll definitely try out.

Ceterum censeo: Raw data now!

by derwisch

Das Ehrenamt, oder Türschloss: 0, siebzehn Feuerwehrleute: 1

Sun, 25 Oct 2009 06:53:51 +0000

Das Ehrenamt

Sie sind der Sauerteig, der die Gesellschaft zusammenhält, und der Kitt, der sie treibt. Sie wirken im Verborgenen und erbringen Leistungen, für die man Profis niemals bezahlen wollte. Manchmal treten sie auch an die Öffentlichkeit und lassen den Leser spüren, wofür sie ihre Freizeit opfern. Ihre Plattform ist das örtliche Mitteilungsblatt der Gemeinde. Ein Beispiel soll verdeutlichen, wohin das führen kann.

Türöffnung, x.y.2009, Einsatznummer: n, Einsatzbeginn: 14:27, Einsatzende: 14:38

In einem Wohnhaus in der Friedrichstraße bestand ein Verdacht auf einen Wasserrohrbruch. Da der Mieter der betroffenen Wohnung nicht öffnete, wurde die Feuerwehr zur Türöffnung gerufen. Nachdem die Tür von der Feuerwehr mittels Spezialwerkzeug geöffnet war, war ersichtlich, dass für die Feuerwehr keine Tätigkeit erforderlich war. Die Einsatzstelle wurde an die Polizei übergeben. Von der Feuerwehr waren 17 Personen und fünf Fahrzeuge im Einsatz. Die Polizei war mit zwei Beamten und einem Streifenwagen vor Ort. Fahrzeugaufstellung: LF 8/6

by derwisch

Finding coarse highways in Openstreetmap

Tue, 30 Dec 2008 20:40:24 +0000

Quality control in Openstreetmap

Openstreetmap is a rich project with diverse users and entries. As such, the data base is rife with insufficient, contradictory or plain ugly entries. Some people would like not to have any entry there rather than one that might be an embarrassment. Others applaud the fact that there is an entry at all, and wait for crowd wisdom to set things straight. And there are others who build tools such as the OSM inspector that hint at possible problems so that helpful gnomes may take care of them.

One of the issues that occurred in the mailing list talk-de was roads that were drawn as coarse polygons, looking like a race track from the Disney movie Tron. Maybe somebody took very few trackpoints, maybe somebody was driving to fast, maybe somebody copied a road from a coarse satellite image or maybe somebody was just a bit lazy.

How to extract coarse highways

General features

Coarse highways are highways that have kinks, serpentines that zigzag instead of bend, long stretches of straight road. They look ugly on the map and are unlikely to reflect the real world. As such, they are characterised by triples of nodes showing

sharp angles
long sides.

Extracting those features

Looking at the tools people wrote to extract features off the OSM database I feel like ten years younger. Many things are written in Perl. My first experiences with XSLT and Saxon suggest that this may even be a wise choice. Perl is relatively fast at extracting information. Especially, a user known as Gary68 has written a Perl module and small sample applications of that tool. So I have written a tool that extracts all nodes belonging to motorway (link), trunk (link), primary, secondary, and tertiary highways and calculates the angle and the lengths of the sides of each triple of subsequent nodes. This Perl script I concocted does simply that. To proceed, do this:

Download the Perl module and save it under the name osm.pm under an appropriate module path in a directory named OSM. On my Linux box, it sits under /usr/local/lib/perl/5.8.7/OSM/osm.pm.
Download the aforementioned Perl script.
Change the two variables in the lines below so that they point to an appropriate .osm file. In the example, I took the OSM files I got from unzipping the contents of mecklenburg-vorpommern.osm.bz2 from the Geofabrik server.

my $osmPath="$ENV{HOME}/osm/data";
my $fileName="mecklenburg-vorpommern.osm";

Run the script and redirect the standard output to a new text file. The file I obtained based on the current download has 145642 lines.

How to look at the coarse highways

General idea

We don't have a good concept of what is wheat and what is chaff yet. Therefore we want to look at the representations, sift through some examples to get a general idea. This is the domain of exploratory data analysis, or dynamic visualisation.

What was exported and why

The script does not exactly compute the angle and the length of the sides of each node triple. Instead, it computes the cosine of the angle and the product of the lengths. I decided to do this because I wanted the information on coarseness represented by two variables, and the sign of the angle (+12° or -12°) does not matter. To have the information concentrated in two variables allows me to scatterplot it. A scatterplot is capable of presenting a lot of visual information to the observer.

Starting the recommended program for exploratory data analysis

I downloaded Mondrian by Martin Theus, a program which is designed for visually analysing large data sets. It seems to be written as a refutation of the prejudice that Java programs have to be slow. The program is simply a *.jar file which can be started by something like java -jar path/to/Mondrian.jar. The program expects the statistical software package R running with the library Rserve. If it doesn't, well, Mondrian will cough a bit and take a bit longer to start. Don't worry about the warnings issued. Then, load the data by using File->Open.

Exploring the data

Firstly, we see that the data type of the IDs has mistakenly be assumed to be numeric, while we'd rather see it being a categorical variable. To change this, highlight the respective variable name and select Options->Switch variable mode. You needn't do that for the nodeID as it takes considerable time to update the mode (did I mention that Mondrian is fast? Wait and see) but it makes a lot of sense to do that for the wayID.

Then, we want to scatterplot the nodes by angle and lengths. Highlight the angle and lengths variables and select Plot->Scatterplot. The well-behaving node triples now sit in the bottom right corner. To get an idea of the concentration of nodes you can press <Left> repeatedly in order to decrease the alpha channel of the dots, but you should then really press <Right> to increase it as we want to single out the offending node triples.

Along the bottom, the sharp angles are seen towards the left. As the cosine has a low gradient for small angles, even small deviations from the right side are to be regarded. Along the right side you see triples with longer sides. In my example, the top node is placed at 11.2 km^2, which means a node where both adjacent edges have an average length of more than three kilometers (Ouch!).

In order to get an idea of where the nodes are on the map, a scatterplot of longitude vs latitude might be a good idea. So we select Plot->Scatterplot on the variables lat and lon. The reulting plot does not resemble Mecklenburg-Vorpommern very well, though. So we have to right-click the plot (or do the equivalent with a one-buttoned mouse) and select "fixed aspect ratio" and "flip axes". Now we can select offending points on the QC scatterplot and see where they roughly are on the map.

But which ways and nodes are affected? Let's highlight the variable "name" and select Plot->Barchart. Do the same thing with the variable "wayID". The resulting screen is quite crowded, but the windows can be placed against each other to allow a good overview even on my non-widescreen monitor. Mark some offending points on the scatterplot and select Sort by->absolute selected. You will get the road names or wayIDs which contain most of the offending points.

For example, the way represented by the ID 28724722 seems to have a problem. Let's select the way in the barchart and zoom in to the highlighted points on the map (using middle mouse button and rubber band).

This looks coarse indeed, but before we write an Openstreetbugs entry, we look at the way in the data base itself.

Apparently, the mapper was aware that the granularity is not up to OSM standards and has attached a FIXME tag already, so we don't really need to jump to Openstreetbugs now.

Conclusion

Quality control can be fun! Can it be done algorithmically instead? One would want to look for several adjacent node triples on a way with long sides and sharp angles. This can be put into an algorithm but a bit more experience and playing around is required. For this, tools like Mondrian (maybe also Spotfire or ggobi) are perfectly suited. Being able to sort by frequency of selected nodes quickly gets us the less than perfect ways at the top. Unfortunately, I haven't succeeded in making nodes directly accessible by a browser yet, so some copying, pasting and typing is required. But then we should look at all cases seperately anyway. After all, in the above example a coarse highway is not the most immediate problem on the map as long as some villages there aren't mapped at all. Some minor gripes remain, for example, the IDs are represented in scientific notation even when they are chosen to be on a categorical scale. But all in all Mondrian may be considered as a graphical quality control tool for several other OSM applications.

by derwisch

Map of Openstreetmap authors in Germany

Sun, 16 Nov 2008 21:58:41 +0000

I am just getting to know the Openstreetmap data and playing around with the most basic elements and attributes. Using the user and coordinate data of all nodes, I set up a graphic that divides Germany into 100 km^2 squares, counts the number of nodes and colours the dots (of size dependent on the number of nodes found here) according to the number of users that contributed at least 50 per cent to the nodes in this square. When you load the SVG you may be able to read the user contributing most in that area by mouseover.

The following observations must be made:

Small red dots at the borders are artifacts. Here the squares contain only small areas and we shouldn't be too surprised to see few nodes and contributing users.
Don't take these statistics seriously. As soon as someone does, you'll find somebody else manipulating them, and they are easily manipulated.
This graphic may not reflect the current mapping activity. Some of the authors may long have abandoned OSM.
The greatest density of nodes is found where people have started to map buildings as areas (Dresden, Gelsenkirchen/Gladbeck). But only few users seem to be that determined, the largest blobs are not the ones with most users.
There are obviously few nodes where few people live: from left to right Emsland, Wendland, Altmark, Uckermark. Few features to map and few users.
Some areas are filled with many nodes by relatively few users. The most striking example is Unterlausitz, but also Magdeburg.
I would expect to see the red dot northeast of Augsburg vanish soon. Busy mappers not too far away.
Is it correct to note that holiday areas (Baltic coast, Allgäu) are mapped by more users than other rural areas?

by derwisch

List of rectified maps

Sun, 28 Sep 2008 07:08:30 +0000

Recently I have discovered Openstreetmap. As I have always loved maps, and have appreciated the approach to free software for so long, I figured this is just the project for me. I am really hooked. Yes, my bicycle tour is still described with Google Maps, but this will change one I have worked out how to embed maps with routes (and until all ways have been mapped; after all, I want to switch to improve).

Metacarta's Map Rectifier has been described elsewhere. On the page there is a list of all recently rectified maps. This list does not show all maps as the page would take too long to build up. As a matter of fact, I don't see where older maps are linked to, I even failed to send a Google Request for maps I rectified myself.

So I put up an ol' and boring, Web 1.0 static HTML link list. May it help me and possibly somebody else to recover maps rectified more than two months ago. Note that this is very crude, not interpreting non-ASCII characters etc. But then it's a crutch for navigation I am missing at the original site.

1 Iran Political Map
2 Concord
3 gefle west
4 Gefle center
5 Gefle East
6 CPKIO Gorkogo
7 (no title)
8 Carlskrona
9 ronneby-stad-1939
10 ostersund
11 tyum14
12 6???
13 ????????? ????????
14 East-France
15 sadfas
16 Panaji
17 HIG
18 Hig
19 hig 2
20 Map of Africa
21 Mirnaya
22 Hanzelijn
23 DLSU-Manila Capus
24 Fridstaplan
25 Fridstaplan
26 Fridstaplan skiss
27 city park gavle
28 Dubai photo from ISS
29 city park gavle west
30 (no title)
31 Casta
32 Viedenska
33 QC Streets
34 S Balls estate plan
35 Valls hage Gefle
36 Forsbacka 1933
37 (no title)
38 Belagenheten af gefle
39 karta ofver staden gefle
40 karta ofver staden gefle
41 karta ofver staden Soderhamn
42 falu stad
43 karta ofver staden Hudiksvall
44 jonkoping
45 kristianstad
46 Wenersborg
47 Bayreuth S?d-Ost B-Plan
48 HE-Ingerkingen
49 Ventspils
50 Ingerkingen - Zweiter Versuch
51 Heggbach
52 Heggbach links
53 Heggbach rechts
54 HN_Wertwiesenpark_S?d
55 Illig_Schraeg
56 bruggacher
57 greenville, sc 1797
58 gertrudenfriedhof
59 Topanga Canyon
60 Will Rogers
61 now without a space in the filename :-)
62 Heggbach Haus Martin
63 Heggbach Haus Regina
64 Stadbezirke Braunschweig
65 Hauptsatzung Braunschweig
66 Brequigny
67 Plozevet
68 Uni Oldenburg Campus Haarentor
69 Oyo
70 Plouzane - Brest - France
71 NSG Okeraue Braunschweig
72 Innenstadt Braunschweig
73 Mandaluyong streets
74 Rohlfsstrasse in Dresden
75 smolenice
76 Hopewell Twp
77 Ojen
78 Ojen2
79 iim bangalore
80 region olomouc
81 Amager Strand
82 South East Isle of Man
83 Elebnispark Heuchelberger Wald
84 Heuchelberger Wald_2
85 skver kalinina
86 Metallurgov
87 Berlin Mitte
88 Severnyuj1
89 Map of Nivala
90 Nvala
91 NSG Riddagsghausen
92 Fachhochschule Gelsenkirchen
93 Zubchaga2
94 Ajaccio
95 mio
96 Braunschweig Museumspark
97 Braunschweig B?rgerpark
98 DK Metallurgov
99 Deichtorplatz 1
100 Deichtorplatz 2
101 Deichtorplatz 3
102 Deichtorplatz 4
103 Hafencity Lohseplatz
104 Brooktorkai
105 Saltlake_kolkata
106 Saint Chads Circus, Birmingham (Oblique)
107 lesopark_60_let
108 Vendome-Trinite
109 Pilvilampi
110 voronezhskie ozera
111 SPB 2015 - 1
112 SPB 2015 - 1
113 Stavba 513
114 Beaverton OR
115 South Caloocan Road Map
116 Shokoladka
117 Braunschweig 331 - Siegfriedviertel
118 Braunschweig "Fl?chennutzungsplan"
119 Lanlaufplan Sulden
120 Haapavesi
121 Braunschweig Botanischer Garten
122 saint anthony park
123 Schrammsteine 1
124 Hamburger Bezirke
125 D?sseldorf Stadtbezirk 10
126 siuslaw
127 Mt. Hood Elevation
128 augsburg
129 auxburg
130 auxburg b17
131 Groix
132 Groix2
133 test
134 NK 53-1 Artem
135 Wels
136 JKU
137 JKU 2
138 olmik prednadrazi
139 Aurora
140 Aurora
141 Aurora3
142 Aurora3
143 hemlingby
144 Samarskaya Luka
145 Beijing-3
146 Beijing-6
147 Baden-Baden Stadtmitte
148 test
149 Mineo
150 Mineo
151 (no title)
152 See Zwenke
153 (no title)
154 (no title)
155 Wuerttemberg1905
156 Pergine Valsugana
157 Unterfranken1905
158 Commenius Moravia
159 SuedWestThueringen1910
160 Schlossarkaden
161 (no title)
162 NordSchwarzwald1910
163 M?nchen-Augsburg1910
164 Century City Precinct Map
165 Thornes Park
166 Lyon 1920
167 (no title)
168 beutlerparkareal
169 Vienna from Donauturm
170 East of Pallas
171 MZJKU
172 hrmbrg_missstrs
173 hrmsbrg-shle_lhh
174 NL outline
175 (no title)
176 South of Mijas
177 lake
178 Clifton, VA
179 Americae Sive Novi Orbis, Nova Descriptio
180 Art Loeb Trail North of Blue Ridge Parkway
181 (no title)
182 Capalaba Park
183 stavba 514
184 Sectional
185 Klingenbergerstr. Nord
186 test
187 IRISA
188 Boracay
189 Discovery Bay
190 test
191 Map 8
192 Centrum Galerie Dresden
193 map 8 rene
194 Rohace
195 (no title)
196 Centro Hist?rico de ?vora
197 Parco dell'Etna
198 Metsakasti
199 Test
200 Waldlehrpfad
201 Skootamatta Lake
202 test
203 map displayed on info sign
204 map displayed on info sign
205 B?ckingen
206 IMAGE-G
207 info sign on historical village
208 Boeckingen Ziegeleipark Sued Ost
209 Atlas - Toubkal
210 Kish
211 Kish Iran
212 phnom penh
213 1
214 two
215 Jaeckleinstra?e
216 Jaeckleinstrasse
217 Gleichen
218 Plan
219 Carleton Arboretum
220 pf Test
221 VI-haut
222 Plan2
223 Schleuse Horkheim
224 puerto montt
225 puerto montt
226 manuel montt
227 antihual
228 antihual
229 G?m hegy
230 antihual2
231 Sontheim00
232 WashU
233 Borodino battle
234 Burma Rapid Response Inundation Map
235 Fez
236 Sontheim Zentrum
237 mytest
238 puerto montt
239 Arusha Tanzania
240 kettering
241 Spring-Oak
242 Oak & Spring
243 Spain-Madrid-La Pedriza
244 Braunschweig University
245 Kingenberg-Nord-West
246 Kingenberg-Nord-West
247 Klingenberg Schloss
248 Prahou na kole - C a Z overview
249 Prahou na kole
250 baunei
251 Forjett Street
252 Forjett Street
253 Moritzburg bei Dresden
254 Boeckingen Sued: Im Jockele
255 Kerest
256 Stadtteile Leverkusen
257 military card stuttgart 1905
258 r?seau routier du Grand-Duch? de Luxembourg Esch Schifflange Bettembourg
259 Little Clacton Essex UK
260 Flers, cadastre bitmap
261 val
262 val2
263 Klingenberg Downtown
264 Little Thurrock
265 Lessines BE
266 wwu1
267 wwu1
268 Grays
269 Winter Hill
270 mestnost1
271 Prahou na kole
272 Twiske South
273 twiske map
274 US
275 Mazaa
276 stromb
277 Tanners Map South Carolina 1833
278 Baltimore City 1869
279 West Baltimore 27c
280 West Baltimore 26c
281 West Baltimore 26d
282 Baltimore City 1895
283 Winter Hill
284 West Baltimore 1876
285 Phoenix
286 ?1
287 Billings
288 Two tree island
289 (no title)
290 Two tree island
291 kettering south
292 twiske9a
293 kt
294 Mecklenburg County 1789
295 hahnweide - germany - BW
296 Corcovado
297 Corcovado
298 Corcovado
299 avernes france
300 twiske1a.jpg
301 Center of Bytom in Poland
302 (no title)
303 (no title)
304 Brugge
305 Andersberg
306 Andersberg
307 Schild Blankenburg
308 athens
309 Schild Huntewiesen
310 twiske3a
311 Dolni Brezany
312 bremen districts
313 Big Rock
314 (no title)
315 Einkaufsland Wechloy
316 NOAA cloud snapshot of eastern europe
317 Schloss Trautenfels
318 Graz_test
319 Duesseltal
320 cr1-test
321 Leichlingen Southwest
322 weh-park
323 gbb
324 beka
325 (no title)
326 beka
327 (no title)
328 Zaanse Schans
329 Max-Eyth-Schule Lageplan
330 schule
331 st
332 lamberti kirche
333 bank
334 weser-ems
335 floetenteich
336 bahnhof
337 streams floetenteich
338 Playa Ancha Chile
339 Montemar 1
340 Montemar 2
341 Montemar 2
342 Montemar 3
343 test
344 Playa Ancha 2
345 rotonda puerto montt
346 sbec
347 tester22
348 H
349 Bahnschule / G?stehaus Regensburg
350 Test Cadastre
351 cadastre beaulieu
352 (no title)
353 montgermont
354 UCT lower campus
355 BPlan Markfeld, Bad Aibling
356 Tyrov
357 Rennes Campus de Beaulieu - cadastre
358 Vesleskarvet - TestRun
359 kleines waldstueck
360 waterplan amsterdam noord
361 beverbaeke
362 Heidelberger Schloss von Koch und Seitz 1891
363 streams waterende
364 wettbewerb
365 SR67
366 SR67-2
367 Englischer Garten, M?nchen
368 Englischer Garten, M?nchen
369 Grabenstetten Flugplatz
370 Nordhausen
371 NDH
372 (no title)
373 N56
374 Tropical Depression 1-E
375 Tropical Depression Alma
376 Land Use in County of Cumberland 1945 (Sydney)
377 Bandung
378 Travis County
379 Chisinau Botanica
380 Studentendorf Schlachtensee
381 DFS m33143bc
382 Flugfeld
383 Eulengreuth Dettingen
384 parc de la tete d'or
385 SK-rivers
386 SK-rivers
387 beverwijk-waterplankaart-1
388 Sulzbach an der Murr
389 Stuttgart
390 SK-rivers 3
391 lauter
392 Karp
393 Karp2
394 Karp2
395 KarpMap
396 Naturpark Sch?nbuch
397 Naturpark Schoenbuch
398 Schloss Pillnitz
399 hartenbosstrand
400 Island
401 playa
402 (no title)
403 (no title)
404 (no title)
405 Universit? di Trieste
406 Military Map CSR - cca 1952
407 test
408 Baustelle A5
409 Sontheim Nord
410 Baustelle S1
411 non
412 Dettelbach_Ort
413 V?gen, Sandnes (Norway)
414 Vaagen, Sandnes (Norway)
415 Avanos
416 Dicescu Kishinev
417 INF
418 border e
419 border e
420 boder s
421 border w
422 road
423 inf grey
424 border n
425 Cadastre etampes
426 cadastre etampes bon sens
427 bfe luft
428 bfe karte
429 swarte moor
430 HN-Sued-FH
431 cadastre etampes Feuille_000_AE_01_corected
432 Hanoi University
433 soltau_krankenhaus
434 test
435 Leingarten Friedhof
436 Z?rich
437 Lambertikirche
438 prueba
439 la choca
440 (no title)
441 (no title)
442 lindenborg
443 USGS Orrs Springs (partial)
444 Leingarten II
445 Leingarten III
446 Trier Euren Wald
447 malacky
448 Trier Wei?hauswald walking
449 ligurian sea
450 NIMA map of Northern Israel
451 Karte_der_Hochwasseraufnahmer?ume_in_Bremen
452 Karte_der_Hochwasseraufnahmer?ume_in_Bremen
453 pl.rybnik old
454 A5 Nord
455 A5 Nord
456 Knoten Eibesbrunn
457 Heilbronn Feuerwehr
458 v
459 Gewerbegebiet
460 pl.zamosc
461 bremen mitte
462 bremen-mitte2
463 Schrames
464 Heilbronn / Zeppelinstrasse
465 HN Fa. Marbach
466 Leingarten Mitte
467 mapa
468 cronulla
469 dupa
470 Bole, Addis Ababa
471 Heilbronn - Viehweide
472 Guides Cosyn Martelange Bodange
473 WABU_I
474 AMD_Dresden
475 NoTitle
476 Giardino storico di VILLA COSULICH - Trieste
477 Parco Farneto - Trieste
478 torbole
479 Zzfahrt Volle Rose
480 torboleII
481 malcesine
482 Old Portsmouth
483 Flap
484 Thaan
485 Oslo
486 wolkersdorfI
487 wolkersdorfII
488 wolkersdorfIII
489 sachsengang
490 Westfriedhof
491 Residential
492 GE Ehrenberg Ost
493 Maribor
494 NorthWestDaejeon
495 Newtown
496 Sapiegine
497 Krankenhaus Dresden Friedrichstadt
498 Sukharava-5
499 Akmola_Province_Kazakhstan
500 Teide National Park
501 Nationalpark Harz
502 BALMA
503 (no title)
504 (no title)
505 Heilbronn Gesundbrunnen
506 Europa
507 Eilenriede
508 Helsinki and Espoo ca. 1900
509 Kirkkonummi, Siuntio ca. 1900
510 Klinikum Oldenburg
511 Kirkkonummi and Espoo ca. 1900
512 FNP Goettingen
513 Gro?raum Braunschweig
514 Stadtteile D?sseldorf
515 Stadtteile D?sseldorf
516 BPlan G?ttingen-Steinsgraben
517 Nationalpark Harz
518 Road Map
519 Landkreise in NW Germany
520 KPN
521 Italy
522 Subcities of Addis Ababa
523 Tube Map
524 Agora
525 Sukharava-5 plan(2)
526 IGA-PARK Nord
527 Bangkok-Suvarnabami Airport
528 Morte, Devon, UK
529 Croyde, Devon, UK
530 Cavusin
531 Wels Traun
532 sdfsdf
533 Lohmuehlwiesen
534 Buch 001
535 Buch 002
536 Buch 003
537 Test
538 1
539 2
540 3
541 4
542 Gurkenland
543 Branitz
544 RHMPLN
545 Borlinghausen1
546 Borlinghausen1
547 Borlinghausen2
548 Borlinghausen4
549 Neuenheerse
550 Bebauungsplan Schriesheim Nord
551 Neuenheerse2
552 Burg Vischering
553 Bebauungsplan Zollverein
554 celle akh
555 Regensburg/Winzer 1
556 Pineta di BARCOLA - Trieste
557 Giardino storico SKABAR di BARCOLA - Trieste
558 krw rog such
559 WBG_LZ
560 Gageneck
561 CONCAT
562 wbg
563 Waldbruecke
564 wbgGF
565 dt
566 Goettingen 1904
567 Heilbronn 1902
568 Schriesheim Altstadt
569 (no title)
570 Google Earth Kumbo
571 Kahren
572 Werben
573 Stradow
574 GraveYard
575 GraveYard II
576 GraveYard III
577 varden1
578 zwierzyniec
579 test
580 varden2
581 (no title)
582 Liegenschaftskarte Cochem und Cond
583 jihocesky kraj
584 vpils
585 Rurtalsperre
586 Oberschaar1
587 Obersch1
588 Obersch2
589 Chisinau
590 (no title)
591 a1 highway so?nica be?k
592 h3ead
593 mere
594 S2 AST S??enbrunn
595 Obs01
596 Obs02
597 Obs03
598 friedberg/germany/1
599 Naturschutzgebiet bei Sandhausen
600 philippine languages
601 GPS1
602 http://geogreif.uni-greifswald.de/geogreif/?cat=31
603 jurmala
604 Zuideinde Oostzaan
605 Leingarten Mitte
606 (no title)
607 S8 ?bersicht
608 B?ckingen
609 Boeckingen-Wasserturm
610 Neubeuern 1855
611 Neuw?hr
612 stationsstraat
613 hermannsburg
614 arziachenna
615 (no title)
616 belle ile 1
617 Syria - Aleppo 01-01
618 308175
619 Boeckingen Sued
620 test
621 Eichbottseen
622 test
623 nike rocket launcher
624 wanfried stadtplan1
625 park_gagarina
626 Hamburg from WP
627 Siebensteinhaeuser
628 Siebensteinhaeuser
629 Liegenschaftskarte Sehl
630 Liegenschaftskarte Cochem Altstadt
631 Liegenschaftskarte Cochem und Cond
632 Liegenschaftskarte Cochem und Cond
633 Wieescher
634 Test
635 P?ttkestour1
636 Ontwerpstructuurplan_Bombraak.png
637 UW Arboretum
638 South collie
639 ederseeruinen
640 Mifflin Township
641 Paettkestour2
642 Paettkestour3
643 Paettkestour4
644 stredocesky kraj - silnice
645 testmap
646 testmap
647 W72
648 Arzachena
649 testmap
650 g?ssenberg
651 test
652 test2
653 (no title)
654 Korfu-City
655 Calgary Cycling Paths
656 Una? - 1
657 Cidade Unai MG
658 AvtoVokzal
659 (no title)
660 Osmecke rmany
661 Osmecke Germany
662 Bahnstadt Heidelberg
663 WIG - Zakopane
664 CIA 1990 Hungary
665 WIG - Krynica (A) 1935r
666 Feldberg-Area, Germany
667 nb
668 nb2
669 nb2
670 RP_MUC_Nord
671 Plan Casa
672 J?terbog_Anlagen
673 Mapa Pueblo
674 Dahouet
675 Zolli
676 Port_Dahouet
677 RNP_Duesseldorf
678 Yalcha
679 tar
680 Gleiwitz 1940
681 wezel sosnica1
682 wezel sosnica 2
683 wezel sosnica 3
684 Yalcha 2
685 Ville Port
686 Ville + Verte
687 (no title)
688 FS_SUED
689 M_Nord
690 Alb-Camping Westerheim
691 Ville Centre
692 kin
693 novgorod_district
694 Ville Piegu
695 Plano Ciudad
696 Plano Pueblo
697 (no title)
698 Ville part1
699 Golf
700 Dresdner Heide
701 Croix
702 Le Cloitre
703 Vissel
704 sn_wil
705 Asel
706 Berlin - Sch?nefeld
707 Berlin - Sch?nefeld
708 Heanor_UK
709 Bergstedt23
710 Bergstedt23
711 bukowe
712 Mosk 18 km
713 Carte
714 wedelschule
715 barrios de ibague
716 Schuetzenbuehlring
717 Ruchheim
718 Ruchheim
719 OC Model
720 Olula del Rio
721 vietnam
722 Toulon Cadastre AO
723 Toulon Cadastre Est
724 FNP Bestwig 3
725 FNP Bestwig 8
726 waldfriedhof ofenerdiek
727 parkfriedhof kreyenbr?ck
728 parkfriedhof kreyenbrueck
729 FNP Bestwig 6
730 Bridge of Allan
731 Bridge of Allan
732 Bridge of Allan
733 Toulon Cadastre CL
734 Waldfriedhof Volksdorf
735 passocorese
736 Farmr
737 farmr detailed
738 Middle Dungeness River 1947
739 (no title)
740 (no title)
741 (no title)
742 gambia
743 Koniak?w
744 (no title)
745 egleny
746 Hwange district
747 Wroclaw, Olesnica
748 (no title)
749 mapa
750 (no title)
751 Wroclaw University of Technology, campus map
752 Wroclaw University of Technology, campus map
753 Zoo Frankfurt Main
754 FNP Bestwig 7
755 Bucknell
756 Trier Maarviertel
757 (no title)
758 New city centre of Lelystad
759 Theodorhof
760 Bplan Langenhorn32
761 Bplan Bahrenfeld7-1
762 Reykjav?k
763 Test
764 Wantage and Grove
765 larran
766 hrm_2
767 wed-11
768 (no title)
769 N?stved Sygehus
770 Test GP
771 New South Wales
772 Fliegerhorst Oldenburg lowres
773 Forli East
774 Caorle East Small
775 Test GPMB
776 Fl?chennutzungsplan Weinheim-Rippenweier
777 Heidelberg Zentrum
778 ENGM
779 Horn
780 Farnborough, Berkshire, UK, 1970
781 Horn
782 (no title)
783 Greenwood 1818 Lancashire
784 (no title)
785 (no title)
786 (no title)
787 (no title)
788 (no title)
789 (no title)
790 Test MSB
791 (no title)
792 (no title)
793 (no title)
794 (no title)
795 Dossenheim Ortsmitte
796 vipava slovenia
797 Heidelberg Gemeindegrenzen
798 (no title)
799 jva kreyenbrueck
800 (no title)

by derwisch

PocketMod and pstops

Sat, 10 May 2008 08:04:47 +0000

The joys of Unix are the versatile tools available. The downside of many of these tools is their usability. A role model for this general observation is pstops. It allows to rearrange Postscript pages according to any order and rotation you like (well, ok, rotation by 90° steps), mostly in order to prepare documents for printing, such as brochures. The versatility comes at a cost, though. First, the notation is a bit terse. I won't go into the details, you can have a look at the man page to see what I mean. Second, adjusting the rotation and shifting requires some geometric thinking. As the rotation is around a corner and happens prior to translation, pages are rotated "out of sight" if not translated, and people lacking discipline in geometric thinking (such as me) are bound to applying trial and error.

Once I have succeeded I am so overjoyed I want to tell the interwebs about it. The application I had in mind is the PocketMod style of folding up a paper printed single-sided to a small 8 side brochure. I tried out the application on the PocketMod homepage but it didn't match my needs. First, you need Flash installed, which is often a bad sign. Then, the stuff you create is ephemeral, at least I have not seen any possibility to download and save your work for later reference. Third, non-ASCII characters are ignored in output. Fourth, the output is, instead of A4 paper, on some local format which I don't have paper for. Fifth, the output is in some pixel graphic format.

Buying into the Unix way to split a tasks into subtasks, I thought about making some Postscript file myself (ok; some PDF file, then converting to Postscript), then applying pstops to rearrange and then to reconvert to PDF. That's one of my other gripes with pstops: That there is no wrapper (I am aware of) for PDF documents. Obviously the tenet is that people who can master pstops don't see any hassle in converting; or they can write their wrappers themselves. Sometimes I want to be pampered though.

Anyway, here is the syntax:

pstops "8:1L@.35(10.5cm,0)+2L@.35(10.5cm,7.425cm)+3L@.35(10.5cm,14.85cm)+4L@.35(10.5cm,22.275cm)+0R@.35(10.5cm,7.425cm)+7R@.35(10.5cm,14.85cm)+6R@.35(10.5cm,22.275cm)+5R@.35(10.5cm,29.70cm)" document.ps document-pocketmod.ps

This is another gripe I have. The pstops manual neglects to tell the users that the arranging specs have to be wrapped in quotes in order to be parsed in bash. Yes, I know that's shell-specific, but I think it wouldn't hurt to mention that in some shells the quotes are required.

by derwisch

43rd birthday tour

Mon, 24 Mar 2008 19:52:54 +0000

On Easter Monday was my birthday (probably the only time in my lifetime).

A neighbour was so generous as to lend me his slick bike and I went on a cycle tour through the snow-covered Odenwald.

There is a small stretch inaccessible by cars (and thus by Google Maps) just north of Hettigenbeuern. So the tour is a bit exaggerated by this detour forced on it by Google Maps.

I took the train back. It was an enjoyable ride, especially as it was to cold and too early in the year for motorcycles, and for grass pollen.

by derwisch

Intercityzuschlag

Sat, 15 Dec 2007 20:27:22 +0000

Früher, als der Intercity noch das Flaggschiff der Bahn war, wurde seine Benutzung mit einem entfernungsunabhängigen Zuschlag belegt. Die Reise mit Fernzügen unterhalb der IC-Gattung war zuschlagfrei. Es gab auch zuschlagfreie Intercitys. Diese hatten häufig Orte wie Munster (Örtze) als Ziel und waren im Kursbuch Schiene mit dem Vermerk "Besonders geeignet für Bundeswehr-Familienheimfahrten" versehen. Diese Formulierung wurde unter anderem von Max Goldt angemessen gewürdigt. Heißt sie doch eigentlich "Besonders ungeeignet für Fahrgäste, die ungern in Gesellschaft von Betrunkenen reisen".

Im Jahre 2002 wurde dann der Interregio faktisch abgeschafft. An seine Stelle trat tariflich der Intercity. Der Zuschlag wurde (vernünftigerweise) entfernungsabhängig. Der letzte fixe Intercityzuschlag, den ich in Erinnerung habe, betrug 7 DM. Die gemeinsame Geschichte von Intercityzuschlägen und Euro war zu kurz, um bei mir Erinnerungen zu hinterlassen. Natürlich hinkte die Umrüstung der Interregiowagen auf Intercitystandard der Tarifumstellung hinterher, aber da die Mehrkosten nicht mehr explizit durch einen eigens zu erwerbenden Zuschlag abgebildet worden sind, sondern im Fahrkartenpreis versteckt wurden, regte sich wenig Unmut gegen diesen Aufpreis ohne Gegenleistung. Vorbei die Zeiten der zuschlagfreien Intercitys, herbei der zuschlagpflichtige Interregio. Krönung war ein intercity, der bis 2004 zwischen Heidelberg und Stuttgart mit druckunertüchtigten Wagen durch die Tunnels der Schnellfahrstrecke im Kraichgau brauste. Es muss wohl mindestens der betriebsärztliche Dienst der Bahn gewesen sein, der dieser ohrenbetäubenden Tortur ein Ende setzte.

Heute ist der entfernungsabhängige Zuschlag beträchtlich geworden. Beispiele wie das nebenstehende zeigen, dass die Verkehrsmittelwahl zwischen Weinheim und Gießen (etwa 140 km) schon mal 11 Euro ausmachen kann, also das Dreifache des Preises vor sechs Jahren. Dies nicht bei dem international bestaunten Geschwindigkeitswunder, das Taunus und Westerwald mit bis zu 300 Sachen fährt, sondern einem Zug mit einem Durchschnittstempo von 75 km pro Stunde.

11 Euro bei 140 km, das war vor gut zwanzig Jahren mal der Grundpreis für die Fahrt. Jetzt ist es der Unterschied zwischen zwei Zugtypen.

by derwisch

Apple treats own messages as spam

Wed, 31 Oct 2007 05:13:46 +0000

This morning I checked my mail. Only two new messages, on of them marked as spam. Upon a second look, this mail was sent by an Apple e-mail address. The mail program shipped by Apple categorized this message from the vendor's domain as spam. Note that I did not go out of my way before to train the filter very much. It seems that the spam filter does not use a whitelist for Apple's own domain. This contributes somehow to the reputation that Apple is more on the side of the consumers, while Microsoft, if pressed, tends to lean to big corporations. Not that I buy into this image, but this behaviour definitely supports it.

Apple marks their own mails as spam. Imagine.

by derwisch

Sealing the Seams

Sat, 11 Aug 2007 13:31:06 +0000

Shaky weather nonwithstanding, we are still planning for a camping vacation. Last time we pitched our tent I found that it starts to show its age. It is a WeatherMaster plus by Coleman purchased in 2002. We have used it since then for a total of maybe seven weeks, during one- or two-week holidays, the two-week periods spent in the Netherlands near the coast. The tent withstood some wind and rain with bravado, overall we're quite happy with it.

The last time we've been using it (a couple of weeks ago), I noticed that the textile tapes which seal the seams between floor and walls from the inside are going off at several places.

It seems that the glue is losing its function. Therefore, I went down to the local outdoor store where they recommended me a tube of seam sealer by Edelmann & Ridder, Isny (it rains a lot down there, so they should know their job). The dealer said that this sealer works on 70 per cent of the tent fabrics being sold, so good luck to me.

This was on Monday. Eager to get things done, I pitched the tent and re-glued the tape on the seam. I put a bit too much tension on it, so I ended up with some spare tissue which forms a small noose now. Then I left the tent pitched to have the sealer dried. The dealer said something like 24 h before it could be safely rolled in without gluing what's not suppost to be held together. The following day it drizzled. The next day also. The next day as well, with intermissions for slightly harder rain.

So went the fourth day. It is not unusual to have four days with occasional rain, but four days of more or less constant drizzle in the mid of summer is uncommon here. The upshot is that we get some proof if our tent still withstands the elements, I thought. And it turned out to be wise to have it undergo such a test. The Weathermaster did not quite live up to its name anymore. Reopening the tent after four days and 32 mm, I found that some water has crept inside and formed puddles.

The puddles were of a size that called rather for a cup and a bucket than a mop.

. I bailed a total of 3 liters. At least the floor is still tight. The tent covers an area of about 10 m², which means that about 99 per cent of the precipitation went into the tent. Given the amount of water the average human sweats out a night, 3 liters does not sound a lot in a four-person tent. Still, even if the dripping fails to hit your face at night, the sound of it can be unnerving enough for this fair-weather camper here. Now on the first dry day (on which the outer walls dried up at an instant), I was putting the seam sealer to good use, paying special attention to the seams above the puddles.

Wish us luck.

by derwisch

Misconceptions about SAS: The MERGE statement

Tue, 07 Aug 2007 16:45:46 +0000

Recently, I used the MERGE statement to attach information about treatment group to the patient data. The patient data have more than one observation (or, to the non-SAS crowd out there, record) per patient while the additional information is patient-specific and therefore has only one observation per patient. Treatment was coded as 0=Placebo and 1=Verum. For the sake of clarity, let's assume the tables were set up like this:

USUBJID	DATE	VALUE	…
1001	20070520	43.8	…
1001	20070521	41.5	…
1001	20070522	39.8	…
1002	20070602	46.2	…
1002	20070603	41.7	…
1002	20070604	43.6	…
…	…	…	…

and

USUBJID	ARMCD
1001	1
1002	0
1003	0
…	…

where ARMCD denotes the treatment group.

Then I did some analysis with repeated measurements per patient. I found out (for the umpteenth time) that SAS, when using categorical variables as exploratory variables, the last category will be used as a reference and all effects are described in comparison to this reference. this meant that Verum was the reference and the relative effect of Placebo was given. This is unintuitive. Not knowing if this behavour could be altered, instead I altered the treatment group assignment in place (Don't ever do that, always create a new variable for this unless you know what you are doing):

data new; merge patients values; by usubjid; armcd = armcd-1;

What does the result look like? For good measure, I already introduced the formatted version of the treatment here:

USUBJID	ARMCD	DATE	VALUE	…
1001	Verum	20070520	43.8	…
1001	Placebo	20070521	41.5	…
1001	Verum	20070522	39.8	…
1002	Placebo	20070602	46.2	…
1002	Verum	20070603	41.7	…
1002	Placebo	20070604	43.6	…
…	…	…	…	…

Yikes! We get different treatment assignments to the same patient! What is happening here?

Time to read the documentation. In the language reference I am forwarded to the tome Language Reference: Concepts. In it you can find the following sentence which describes what the system does when match-merging:

SAS retains the values of all variables in the program data vector except those variables that were created by the DATA step; SAS sets those values to missing.

The operational word is "retains", or shall I say RETAINs?

SAS mostly works on data stored in tables; like Perl, it is very good to read and process one line at a time. Very few actions happen between lines or carry information from one line to the other. This is part of what makes SAS the number-crunching behemoth who is not afraid of these 20 million observations of telecommunication customers. This behaviour makes the system inflexible on the other hand, yet it is ingrained in so many users' minds (such as mine) that we act surprised when the system actually uses information from one line in the next one.

When programming, I had the following model in mind: "I assemble a new table by merging one table from the left, the other from the right. There is one variable in the new table which I would like to recode in place." The erroneous assumption is in the words "in the new table". Instead, SAS retains a copy of the variables to be merged in its internal memory until the last line of merge value matches has been reached in both data sets. This internal memory is called "program data vector". All operations are happening on the variable in the internal memory now. If the code armcd = 1 - armcd; is applied to it, the value of the variable will merrily hop to and fro, between 1 and 0, until the next USUBJID is reached, in which case it is set to the correct treatment group of the next patient (whose data lines will be accompanied with zebra stripes of toggling 1 and 0 values again).

Moral: Don't ever overwrite variables in place, create new variables instead.

by derwisch

Forest plots

Tue, 07 Aug 2007 16:36:21 +0000

A couple of weeks ago I sat in a SAS course to refresh my memory about graphics in SAS. Near the end of the talk the question was posed if SAS could do forest plots. The speaker, not knowing the term, asked for an explanation.

A forest plot is an arrangement of bars which represent clinical trials or other studies roughly addressing the same research question. The ends of the bars represent the (95 per cent, of course) confidence limits associated with the estimated effect. At the point estimate a square is drawn. Its area represents the weight given to the respective study in the elicitation of a common effect estimate as part of a meta analysis. The common effect estimate is represented by a diamond whose width corresponds to the confidence interval of the pooled estimate.

I had to admit to myself (and, as of now, to you) that I was unaware of the term forest plot too. I am well aware of the funnel plot, where the bars are horizontal and arranged on the y axis according to the number of observations in the study, or the standard error (in which case the y axis points downwards).

The idea is to get an idea of publication bias. If anything like that exists, weak studies with few patients are more prone to fall prey to it, leading to an asymmetric picture, looking unlike an upturned funnel. My first reaction was "a rose by any other name" and that a forest plot does not really convey anything new. Both funnel and forest plots stem from the Evidence Based Medicine pundits. The funnel plot seems to be a bit older; the first Medline search result is from 1988. The first search results for "forest plot" in Medline carry titles like "Biological screening of rain forest plot trees from Palawan Island"; I haven't found a citation relevant to this subject from the last century at all.

Maybe the choice of the descriptive tool is a way of forgoing the conclusion: The main task of a funnel plot seems to be to expose rather to investigate some publication bias. If your goal is to distill a common estimate from different results, you'd better not undermine your results with a tool designed for that goal. Rather, you will use the forest plot as a more neutral means. But maybe I am being cynical here.

First of all, both the forest plot and the funnel plot reveal in two graphical features what is essentially the same feature of the underlying study. The weight of the study is, of course, deeply linked to the width of the confidence interval. This holds ex forteriori for the standard error; yet also to the sample size. Therefore, some versions of the funnel plot only plot the point estimates as dots, hiding nothing. Yet the urge in properly educated statisticians is strong to pay the confidence interval due respect and never only plot the point estimate. Giving the confidence interval implicitly by some position on the y axis just doesn't cut it.

Once the confidence intervals are part of the plot, another issue arises. The weaker trials are rendered by wider bars "and, perversely, were the most noticeable on the plots". Therefore, in the desire to appeal to intuition rather than to sense of abstraction, some way of stressing the short bars had to be found. Enter squares.

Anybody who has done plots where a measure is represented by the size of a figure knows the problem, as does anybody who has ever populated an aquarium without putting too much thought into foraging behaviour: The larger objects will swallow the smaller ones. In this case, the problem is virtually by design: The shortest bars will be covered by the biggest squares. If you choose the squares to be small enough you might have another problem, especially with studies of varying size: The smaller objects will not be larger than the hairline width of the bar so you will have to find a way to make the location of the point estimate noticeable. Some other problem is that the squares might overlap other bars, or even each other. Somebody came up with the brilliant idea (PDF) to replace squares by rectangles of fixed width. Now we get bar-swallowing blotches that grow linearly with size of the study, and not just the square root. To give the Evidence Based Medicine community credit: This variant didn't gain popularity.

Once the squares were introduced, it wasn't felt necessary to stress the, albeit important, feature of the size of a study by three graphical features. Exit the y axis. Enter equidistant bars.

One ambiguity in forest plots is whether to draw the bars vertically (PDF) or horizontally. The vertical arrangement gives some justification to the name. The horizontal version reminds me (and, I presume, my cohort-mates) of a scene of the arcade game Frogger where the logs are floating in a river and used as stepping stones. Indeed, it should have been called Frogger plot to avoid connotations with completely unrelated methods (and unrelated Medline search results to boot, see above). Stuff SEGA and their trademark rights, this is freedom of research. As legend has it, they can't even agree on the spelling of for(r)est plots. As in 1990's grrls, the number of r's is optional. The horizontal arrangement has the immense bonus that single studies can be labelled. And, remember, all you Tufte-consuming statisticians, always value labels over legends and footnotes.

Now we have grudgingly accepted that both square size and confidence interval bars are part of the plot, and that the bars are arranged in an equidistant fashion, we gain one more degree of freedom to convey information, that is the order of the bars. It should not be wasted by ordering them arbitrarily. So what should the order convey?

You can order by size of the trial, in which case you re-introduce the funnel plot in a distribution-free version. But we already told the reader by two features how precise the trials are, thank you. A more sensible possibility is to order by point estimate, as done in another SAS macro (PDF) (I hope I have helped answering the initial question).

I like this version; it gives you immediately the idea of dependence between precision and point estimate, what the funnel plot is intended to show. Then again, this arrangement does not add any new information but only stresses what is already said. More convincing to me would be the order of publication dates.

The statistical approach to meta-analysis is essentially to treat a single study as a draw from a pool of independent objects. Anyone who has ever undergone the painstaking process of a study, be it a multicentre clinical trial or a case-control study in quest for a cause of a disease, will have a hard time to accept it being treated as just one of some interchangeable objects drawn from an urn. Obviously the studies are not independent as they (hopefully) cite each other. A new study about a research question addressed by a previous study is conducted for a reason. Take my words with the contents of a salt cellar as I am speaking as an outsider here, never having conducted a meta-analysis. But it is precisely for that reason: Although the aim to aggregate a common estimate is laudable (she who claims you couldn't do it because every study is unique questions external validity of any study), I feel that the statistician's role is one of the less important in the endeavour.

Therefore, any information which sets the single study apart should be conveyed in the plot. Labelling the bar with author and year, but also with peculiarities of the study among the others (slightly divergent hypothesis, choice of collective …) adds crucial information to the graph. As in Frogger, it is important to notice the alligators among the floating logs — the studies you should rather cautiously use as stepping stones. If space allows, one may even plot the mutual citations between the study publications as a (hopefully acyclic) directed graph next to the forest plot.

To sum it up, there are some upsides of forest plots over funnel plots. Funnel plots have an advantage where there are many studies which are better represented by just one dot. If there are fewer studies, forest plot them, but take notice of the individual trial as much as space allows.

by derwisch