It’s a hack, but it’ll work. You’ll need either:
- Windows, running the git-bash app (highly recommended). Download it here: Git – Downloads
Bitcoin-Core version 12.0.0 shipped with a new heavily optimized elliptic curve math library. Prior to the release, there were whispers that the new library would significantly speed up the blockchain sync – a sync which was taking me 12-24 hours to complete. Excited for the potential speed up, I decided to benchmark the sync and talk about it on reddit live. Needless to say I was amazed at the difference. Thanks to the core developers for this!
How long does it take a new Bitcoin full node to sync to the blockchain? Does the recently merged libsecp256k1 speed up the sync?
These are good questions, so I wrote a small bash script that gives us some data about the sync time of the new release candidate, whose new optimized library for elliptic curve operations has been rumored to give up to a “700%” increase in the speed of verifying transactions. Disclaimer: it’s not wonderfully modular or well thought out; I did it quickly just to see how the performance was.
So let’s benchmark bitcoin-core with the new libsecp256k1 to see this speedup ourselves. Here is a script for making a log and getting a birds eye view of the sync process. Spoilers: it was really, really fast.
This is the kind of thing I always thought about doing, but never did. It’s a simple thing to do, however it just takes time to write up a useful albeit simple script to give you the data you want. Not only that, but previously, when syncs were taking days to complete, this sort of test would eat up my machine.
Since I have been involved with bitcoin, I have known the user base to be (I love you guys) …well, prone to getting carried about with the fear of an impending doom of the BTC — especially if the inevitable doom will be brought upon by some physical limitations of the client, its capacity, and scalability.
In fact, back in 2012-2013, the total size of the blockchain was really all the rage, and the buzzword of the times was “blockchain pruning”. This involved lot of media focus on this problem, as well as efforts by developers and theorists alike to put forth a concerted effort to find a method for reduction of the very scary, problematic growth rate of the block chain!
That’s right everyone was talking about the block chain being too large on disk already, and growing too fast. How bloody ironic, eh. I won’t say it. Don’t even say it!
Over time I have built up a thick skin for these sorts of mobs of fear and uncertainty. I just keep using my coins and really tend to take the doomsday scenarios with less than a grain of salt, because I have seen bitcoin overcome obstacle after obstacle, so many times that it becomes almost illogical to be anything but calm through it all, and feel as though we are in good hands with the devlopers (we are).
But node sync time has been a subtle concern to me for many months – not because I read it on www.scary-bitcoin-news.ninja, but because I noticed this sync time change over a long period of time and become a real hassle to me and running my node as a casual.
In early 2013, it took “pretty damn long” to bootstrap a new node. But I was able to run a full node that year, all year, on a mid level laptop, that I would wipe and re sync every few months. But by the end of 2014 I was no longer running the laptop node and noticed long sync times on my main high performance monster of a desktop. I felt it took so much longer in fact, that I considered not even running the node, which would have really sucked for me, and put a big damper in the way I got used to used bitcoin. By the end of 2015 I was just trying to avoid a fresh node sync at all costs, which is hard as a man who re-installs his OS every few months. The sync would, on either of my computers, take between 36 and 96 hours for me, depending on how uninterrupted I could run the process and the state of my internet service, and other variables.
It dawned on me at some point that if I was dreading a sync, I couldn’t have been the only one. And so I did consider whether node sync time was a real deterrent to running a node.
However I decided to go for it because I thought if the figures I heard are true, and this library gives a significant speedup, it might require less than a day. And when I inquired in the IRC channel, I was told “could be as fast as 4 hours.” Hot damn – this I had to so. So I finally did it. Result: Sp happy! It finished in a blazing 4 hrs, 4 minutes, and 4 seconds! Sweet victory for node operators and bitcoin as a whole!
This did truly make me very happy as it served my purposes and the community as a whole in no small part. I was stoked, and to such a degree that had to spread the joy around the community, encouraging others to take note of the speedy sync time by benching it themselves. I made a reddit post, and put the code on github, along with the data from my 4 hour full bootstrapping of bitcoin core.
Well anyway, I enjoyed this little treat from the devs, and hope y’all do as well. Thanks guys! Long live the core!
Messing with LS_COLORS
Before and after
I’ll first set
LS_COLOR='' Below, you’ll see that nothing is colored, except for the directories which are blue, because of another environment variable called DIR_COLORS. We are going to learn how to tell LS_COLORS to colorize files by name.
I usually had colored by files by whether or not they were executable, a symbolic link, etc, which something like
LS_COLORS='di=96:fi=0:ln=33:ex=1;92' which sets
- directories to 96 (cyan)
- files to 0 (default)
- links to 33 (yellow)
- executables to 92 (green)
and so on.
I realized recently you can set LS_COLORS to colorize files by name. This is useful in times where you know many types of files will be present in a directory and you want to see what they are right away. Or even if its not that useful, it’s still pretty cool.
Here I know I have .tex, .aux, .log, .bib, .png, .zip, and .pdf all in this directory. I do
Notice how you can set bold and a color with 1;COLOR.
Since you can take this to extremes, I did. I quite like it – I use colored pens in real life, so when I discovered this, I just had to …
🙂 Enjoy the pretty prompts!
Get bitcoin price data from the coinbase API and parse it into an easily manipulated form.
>>> from fast_dump_v22 import *
Will return all the data on page 1 of coinbase’s bitcoin price API (this is the latest data).
You can almost always turn on optional debug statements
>>> get_first_N(3, show=True)
... fetching first 3
You can get all the price data that coinbase has
... getting page 1
... getting page 2
... getting page 3
... getting page 4
... getting page 5
... getting page 6
... getting page 7
... getting page 8
... getting page 9
... etc ...
All the ‘get_*’ functions return a price_data string, which is interlaced timestamps and prices littered with newlines and commas. You can print them to see what is going on more clearly:
Turn on the optional show switch for printing large vectors
>>> prices(get_page(11), show=True)
... returning 11000 prices in specified range ...
You can use prices(data)[k] and timestamps()[j] to return the kth price in data, or the jth timestamp in data.
>>> data = get_page(1)
>>> prices(data, index=4)
are two equivalent ways of returning only the 4th price in the requested range (in this case, page 1). This also works for timestamps.
>>> timestamps(get_page(1)+get_page(2)) == timestamps(get_first_N(2), index=1166)
This shows the expressiveness of this module. In general:
>>> prices(get_page(2)) == parse(get_page(2))
prices() and timestamps() are just functions that return a parsed() object having a specific index, or indices.
>>> parse(get_page(1)+get_page(2)+get_page(3)) == prices(get_first_N(3))
>>> parse(get_page(2)+get_page(3)) == prices(get_range(2,3))
The parse() function is there to manually control the outputs instead of just getting prices, or timestamps
>>> x = parse(get_page(1))
As you can see, parse(price_data)[k] returns the kth price in the list. Indices [k] return the kth timestamp.
The parse() function takes care of some weird edge cases:
>>> get_first_N(3) == get_page(1)+get_page(2)+get_page(3)
>>> parse(get_first_N(3)) == parse(get_page(1)+get_page(2)+get_page(3))
>>> x = get_page(1)
>>> y = get_range(2,7)
>>> prices(get_first_N(7)) == prices(x+y)
OPERATOR( get_page(1) + get_page(2) + ... + get_page(k) ) == OPERATOR(get_first_N(k))
OPERATOR is parsed(), prices(), or timestamps(). We also know prices() can obviously display and return ranges of values. When returning large vectors, you can verify their length by setting show=True. The “show” parameter is optional for all get_* functions and provides some information about the operation being performed.
>>> print( prices(get_first_N(11), show=True) )
... returning 11000 prices in specified range...
since each page is a thousand pairs of values (timestamp, price).
>>> prices(get_page(2)) # returns a long list
fast_dump_v1* are older versions that are somewhat different. They are designed to store the fetched data in the .data directory. This in v2*, this was abandoned in favor of stdout redirection.
// 2015 altodinerd.com
A brief introduction to getting started with bitcoin on a linux distribution
For the impatient: Skip to step 1
Bitcoin is one of the most interesting technologies we have seen emerge in the past ten years. And it confuses everyone who sets out to use it. This post is my definitive guide to beginning your journey with bitcoin the right way.
To me, the right way includes, but is not limited to:
- Running bitcoin on a linux machine. My reasons for this are too many to really list here, but suffice it to say that since linux is widely regarded as the most secure operating system available, and the usage of bitcoin involves being your own bank, using bitcoin on linux is the best way to ensure your coins stay safe.
- Using the bitcoin-core wallet software. There are other wallets available. You actually don’t even need a wallet software to own bitcoins or even spend bitcoins. However, if you want to understand how bitcoin works under the hood – which you should, because you are going to be your own bank, you should use this software.
- Using the command line to do bitcoin things. Actually, using both the GUI and the command line is the way to go. This tutorial (this is part 1 of 2) will be a command-line-centric description because afterall, I’m trying to show you the right way to do things.
But that’s just, like, your opinion man.
It is. This is my blog.
You shouldn’t necessarily trust me or anyone – bitcoin isn’t about trust. But if you google “altoidnerd”, you’ll find that I have been around the bitcoin eco-system for quite some time. That’s all I can tell you.
There are just so many wallets available – why bitcoin-core? Because I have been using it for years, I have a very good method of keeping my coins safe, and that is what I am sharing with you today. Use it- it is THE wallet.
But altoidnerd, I don’t want to run linux. I like windows / OSX
I am a linux and an OSX user. So please understand the scope of my knowledge includes these operating systems. If you want to use OSX, most of this tutorial will make sense with minor changes perhaps.
If you want to use windows, that’s ok, I’m not going to judge you. Bitcoin-core is available for windows as well. Please understand however, I am not a windows user, so I will not be able to give you “the definitive bitcoin windows guide”. You can still read on, however, since you can apply much of this tutorial to usage of bitcoin-qt’s debug window instead of the command line.
Step 1: Download and install bitcoin-core
Navigate to bitcoin.org’s official download page and select the bitcoin-core distribution for your operating system. And do choose linux (tgz). Do that because you should be running linux. But if you aren’t, even though you should be, choose the distribution for your operating system. Extract and install the client.
The “right way”
To really take control of things, and understand what is happening, you’re going to want to run bitcoin software from the command line. Though it isn’t really necessary to download the thing from the command line, I’m going to describe that here because this tutorial is the first installment of a start to finish command line approach. For a sneak peak of what I mean, see an earlier post where I described how to use bitcoin-core to generate QR codes without having to trust shady ass websites.
Go to your home directory and create a new directory just for bitcoin. This is not a necessary step, but once again, this is my “right way” tutorial and I will explain reasons for this later on. Enter the directory, and download the tarball with the “wget” command. The extract the tarball like so:
tar -xzvf bitcoin-0.11.0-linux64.tar.gz
Sweet. Now change directory to where the binaries are,
and launch “bitcoin-qt” (qt means its the GUI version).
It’s going to show you the dialogue box below and ask you if you’d like to use the default directory. Do it. Click “OK”.
You’ll see a friendly startup screen if you’ve done things right!
Step 2: Encrypt your wallet, and wait a few days
Bitcoin-core is called a “full-node” implementation, which means the first time you start it on your machine, it’s going to download every single bitcoin transaction that has ever taken place. Wait until the blockchain is synced to start making transactions.
In the mean time, you need to set a passphrase for your wallet. This is super important. In the upper left corner of bitcoin-qt, go to settings -> Encrypt Wallet, and set a very strong passphrase.
This will make it impossible for anyone to send bitcoins from your wallet without entering your password:
Just make sure you:
- Make your password extremely long. I don’t know if there is a length limit in the bitcoin-core code or not. Regardless, seriously push your personal boundaries on what you think a safe password is. Make it super duper long, and impossible to guess. Wise men have discussed password strength to a great extent. Pick a strong password. 20+ characters. Go for 30. No 40. Just make it long, and keep it an absolute secret.
- Never, ever lose or forget this password. If you lose or forgot your password, your coins will be unspendable forever. Neither you, nor anybody on earth, nor God himself will be able to spend your coins. Your coins will be effectively lost.
Picking a strong passphrase makes security so, so easy
Once you encrypt your wallet, you can copy and store
wallet.dat everywhere (I explain
wallet.dat in great detail in “Step 3”). I literally have copies of my main
wallet.dat on 6 or 7 computers, because come hell or highwater, I will always be able to find a copy of my
wallet.dat file. I suggest you do the same. Here is why strong encryption and insane redundancy works:
- If your password is ridiculously long, you can be very lazy and downright cavelier about where you keep your
wallet.dat, because even if someone finds the wallet file, they wont be able to do a damn thing with it. They won’t be able to spend your coins without your passphrase.
- This means that you can make copies of
wallet.datand store them like literally everywhere, on all your machines, three times over. In the cloud. On your mom’s computer. Hell, theoretically speaking, you can post your encrypted
wallet.daton the internet and just about everyone will have a copy, and be unable to spend your coins, because they don’t know your password. Someone remind me to put some coins into an address, and post the encrypted wallet online to prove my point.
If you encrypt your wallet like a boss, you never forget your password, you never tell anyone your password, and you copy the file to everywhere you possibly can imagine, you will never lose your coins. And nobody will be able to steal them from you. Seriously, nobody. It is computationally impossible.
Step 3: Understand what you just did on your computer
Let’s break it down this way. Here are some things that just happened when you launched bitcoin-qt for the first time:
- A hidden directory
~/.bitcoinwas created. It contains among other things the single most important thing ever: a file called
wallet.dat. This file is critical.
~/.bitcoin/wallet.datcontains all of your private keys.
wallet.datis life. When you see in headlines in the news like “Frustrated gentleman quite upset having lost a usb jump drive containing 100,000 BTC“, it means he doesn’t have a copy of
wallet.dat, so neither he, nor anyone on earth, nor God himself can spend his coins anymore.
- Your computer started downloading the blockchain. The blockchain is a record of every bitcoin transaction ever to have taken place. This will take a few days. As of today, the blockchain is about 42 gigabytes. Make sure you have space for it.
Do not delete your
wallet.datfile. Just don’t. If you want to start a new wallet, instead of deleting
wallet.dat, just rename it to something else, like
wallet.dat.old, restart bitcoin-qt, and it will create a new wallet for you. Keep your wallets people. Keep them good.
Notice how I have a file called
wallet.dat.default? That’s because I never delete a wallet. When I installed bitcoin on this computer, I moved the
wallet.datthat stores my coins into
~/.bitcoin/, but first I renamed the existing file so I can keep it. Because, why not? I’m telling you – don’t delete a file called
wallet.dat, because mistakes happen, and its better to have hundreds of files named
wallet.dat.*than it is to lose your coins.
- A hidden directory
In my next post, I will describe some of the features of the bitcoin-qt debug window,
bitcoind and its helper
bitcoin-cli. You can do interesting things, like import new keys, or convert keys to qr-codes.
If you have ay questions, please feel free to comment or contact me.
If you liked this post, you can donate bitcoins to me here: 12gKRdrz7yy7erg5apUvSRGemypTUvBRuJ
Learn about the author here:
On encoding private keys from Bitcoin-core as QR codes for use in paper wallets, other wallet software, etc …
Note: Breadwallet will not import the private key upon scanning the qr code. It will offer to transfer the funds in that address to your breadwallet keyring.
Breadwallet, an awesome open source iOS bitcoin SPV client, allows you to import a private key for use within breadwallet. So, if you’re a user of bitcoin-core, and want to be able to spend some of your funds you have there with breadwallet, you’ll need to create a QR code of your private key to scan it with breadwallet.
This tutorial uses qrencode, an open source string to QR converter. The source can be found here: https://github.com/fukuchi/libqrencode/
Alternatively, you can install qrencode on ubuntu/debian through repositories
sudo apt-get install qrencode
Then, as of bitcoin-core version 0.11.0 you’ll need to run
bitcoind rather than the GUI
bitcoin-qt. If you’re using an old version of bitcoin-core that does not have the
bitcoin-cli tool, you can get the latest version of bitcoin-core and extract it without having to verify the blockchain again or mess with wallet.dat – your
~/.bitcoin directory will remain unchanged.
Get the newest version of Bitcoin-core
tar -xzvf bitcoin-0.11.0-linux64.tar.gz
You should see the executables for the bitcoin-core suite here.
Run bitcoind and create your QR
Be sure you quit any versions of bitoin-qt or bitcoind that are already running, then start the bitcoin daemon
The & operator makes sure bitcoind detaches from your terminal so you can keep going.
Now you are ready to get a QR code for your desired address. Is your address is “addr”, you can generate a QR code of your private key without ever showing your private key on screen in plain text. Just do:
./bitcoin-cli dumpprivkey "addr" | qrencode -o ~/Desktop/Key.png
Or better yet, if you want to create a new address and QR-encode its private key,
./bitcoin-cli dumpprivkey $addr| qrencode -o
echo "private key dumped for: "$addr
The QR code should appear on your desktop. Below is a screen shot of me doing an example where I create a new address, and get a QR for its private key. Enjoy!
Note that if your wallet is password protected (which it really should be), you’ll need to first do
./bitcoin-cli walletpassphrase "your_password" 100
where the argument 100 means the wallet will allow all actions (like spending, or dumping keys) for 100 seconds. That should be plenty of time.
I know Vi Hart discussed it, but she may have not been the first to entertain the idea. Not sure. The question is do they digits of pi contain copies of hamlet?
Yes, yes they do. Not just copies of Hamlet, but actually copies of every book imaginable are contained within the digits of pi an infinite number of times.
Using software package pi and simple shell commands,
I found that any N digit sequence probably appears within the first 10^N digits of pi ; moreover, the sequence would appear about 10 times in the first 10^(N+1) digits of pi, 100 times in the first 10^(N+2) digits, etc.
For example, lets say our sequence is the last 4 digits of my phone number – 1345
That’s 4 digits, so if I scan the first 104 digits of pi, sure enough, my phone number occurs 1 time. If I scan 105 digits, it occurs 12 times. For 109 digits, it occures like 98 times. Its clockwork. See below.
altoidnerd@HADRON:~$ pi 10000|grep -o 1345|grep -c 1345
altoidnerd@HADRON:~$ pi 100000|grep -o 1345|grep -c 1345
altoidnerd@HADRON:~$ pi 1000000|grep -o 1345|grep -c 1345
altoidnerd@HADRON:~$ pi 10000000|grep -o 1345|grep -c 1345
Anyway, so the question is how many digits of pi would we need to search through to likely find about 1 or 2 copies of hamlet in the sequence?
So I got hamlet in plaintext from 2 sources. I found that hamlet when compressed from plain text to a tar.gz archive, the average size was 69 KiB. Therefore, (I think?) that means to represent hamlet as just a binary integer, it would have 69 x 8 x 1024 digits = 565248 digits.
To get the number of digits in base 10, we multiply by log_10 (2) +~ .69 so the base ten hamlet would be like a sequence of about ~390,000 digitsm (0-9).
Ok so back to pi. How many digits of pi must we scan to probably find a string of length 390,000? That would be 10^390000 digits. That’s a lot of pi, but pi has got enough digits to spare. We should see approximately 1 copy of hamlet in the first 10^390000 digits of pi.
Even cooler is that if we just increase the power by 1, we should see 10 hamlets; increase the power by 2, and we should get 100 copies of hamlet. So quickly, we end up with infinite hamlets in the digits of pi. And not only hamlet … this argument should would for any book.
 This is not too surprising since pi is believed to be a normal number, though this is unproven.
One important difference between NQR and NMR is that for NQR, the transition frequency is site specific and cannot be chosen by the experimenter. In NMR studies, the excitation frequency of the nucleus is just the gyromagnetic ratio times the applied field. Thus, the difficulty in NMR is controlling the uniformity of the applied static B. However, the experimenter is free to choose an operating frequency by adjusting B. For NQR, the transition frequency is proportional to the electric field gradient (EFG) at the nuclear site, which is entirely a property of the substance and a function of temperature. While it is theoretically possible to apply a prescribed EFG in the lab, it is unfeasible as the EFG’s in crystal strcutures are often on the order of many kV/m^2. Thus the particular difficulty in NQR spectroscopy is locating the resonance in the first place, which is much like combing the desert if the apparatus runs at a fixed frequency. The home made pulsed superheterodyne spectrometer used in the trial experiments has proven to have excellent resolution. However, it has essentially no ability to perform frequency sweeps as is done in continuous wave methods. The operator must choose a relatively narrow band (< 1 MHz) to target, and design a custom bridge, tank, and receiver configuration just for small band. Band switching is non-trivial with high resolution pulsed techniques.
At one time, it was not possible to calculate the EFG for all but the simplest of structures. However, DFT calculations have become very potent and accessible in the last decade. Beginning in June of 2013, Calculations of the EFG tensor in dichloro-tetrakis-thiourea-nickel NiCl_2 [SC(NH_2)_2]_4 were made using the GIPAW method with closed source CASTEP software (by then graduatre student Dr. Tim Green at Oxford University, with advisor Dr. Jonathan = Yates). Then in October of 2013, Dr. Ari Seitsonen and I repeated the calculation at ETH in Zurich using the open source DFT-GIPAW package Quantum Espresso (QE).
The two calculations show good agreement in many respects and indicate that the 35Cl resonance frequencies in DTN are likely about an order of magnitude lower than initially believed. Because chlorine NQR transitions occurred around 30 MHz in both para-dichlorobenze and NaClO_3, DTN was initally thought to be likely to show 35Cl NQR in the 29-31 MHz band. This was tested first by Robert Baker (REU Student, 2010) and then in 2012-2013 by myself to no avail.
There is no particular reason 35Cl should have to have a resonance in this 30 MHz band. 35Cl NQR measurements have been made in many materials below 18 MHz and above 60 MHz. A high degree of symmetry in a given structure can result in very low NQR frequencies, with correspondingly very poor S/N. Though all nuclei with spin S >= 1 are guaranteed to have at least one non-zero NQR transition frequency, NQR is not always found. This is likely due to poor S/N when the EFG is small and the transition frequency is very low.
Evidence for the validity of the DFT calculations is that the prediction for NaClO3 35Cl NQR matches observation remarkably well. Additionally, since both DFT packages output the full EFG tensor, the NQR transitions for 14N were also predicted by the same calculations. There are sixteen nitrogen nuclei in each unit cell of DTN (right). DTN contains 4 independent thiourea molecules in each unit cell, each having 4 nitrogen centers.
The NQR frequencies calculated by CASTEP and QE for 14N in DTN match very well with those observed in pure thiourea from literature (David H. Smith and R. M. Cotts). The paper by Smith and Cotts quotes NQR in 14N in pure thiourea NQR at 2.6 MHz and 2.0 MHz for inequivalent sites at room temperature. In the absense of any DFT calculation, the 14N resonance in pure thourea would be a best first guess as to the NQR transition for 14N in DTN.
As the NQR frequency at a given nuclear site is directly proportional to Vzz (the electric field gradient, or stretching), crystals with a high degree of symmetry will have low NQR frequencies. Inspection of the crystal structure superficially indicates a widely symmetric nuclear distribution about the chlorine sites, as the thiourea groups are distributed symmetrically and the structure parameters show each Cl is roughly equidistant to both adjacent Ni nuclei.
The quadrupole coupling constant and NQR transitions The quadrupole coupling constant Cq is defined as Cq = e*Vzz*Q/h  where Vzz is the largest absolute eigenvalue, e is the electron charge, Q is the quadrupole moment and h is planck's constant. We consider a nucleus of spin S and define A = eVzzQ / (4S(2S-1))  where S is spin and Q is the quadrupole moment. Then the NQR frequency or (frequencies) are given by f_nqr = 3|A| / h (2 |m| +1)  where m is the lowest of the two levels m and m+1 over which a transition has occurred. For integral spins there are S unique transitions. For half integral spins there are S – ½ unique transitions. NQR is observed in nuclei with I=3/2, I=5/2 and I=7/2. For I=1/2., there are no transitions. The I=5/2 and I=7/2 are indeed complicated, but for I=3/2 there is a degeneracy that causes there to be one frequency of half the quadrupole coupling constant, which makes the single transition frequency in the case of I=3/2 equal to one half the quadrupole coupling constant Cq = eVzzQ/h = e^2Qq/h. NQR Frequencies for half-integral spins Chlorine has I = 3/2 so there is only 1 transition. Using I=3/2 in the formula above, we obtain f_nqr = (1/2) eVzzQ/h = (1/2) Cq  for axially symmetric field gradients. For non-axially symmetryic gradients, we define the assymetry parameter η = (Vxx-Vyy)/Vzz. η=0 in the axially symmetric case, where Vzz is the only nonzero component. When η is non-zero, the transitions for I=3/2 are f_nqr ~ (1/2) eVzzQ/h * (1 + 1/3 η^2)^(1/2)  In general for half integral spin of I = n/2 n 3, 5, 7,... use what is given in Hahn and you should find the coefficients are 3/10, and 3/20 for I=5/2 f_nqr_[5/2 --> 3/2] = (3/10) eVzzQ/h  f_nqr_[3/2->1/2] = (3/20) eVzzQ/h  NQR frequencies for integral spins For integral spins, we use the formula  again. For for I=1 and an axially symmetric field gradient (η=0) , the NQR transition frequency is f_nqr = (3/4) eVzzQ/h = (3/4) Cq  EFG->C: https://github.com/tfgg/magres-format/blob/master/magres/constants. CASTEP EFG DATA (DTN – 35Cl) relaxed Cl 1 Cq: 8.4298 (MHz) Eta: 0.1272 Cl 2 Cq: 0.1705 (MHz) Eta: 0.5898 Cl 3 Cq: 0.1866 (MHz) Eta: 0.5340 Cl 4 Cq: 8.4513 (MHz) Eta: 0.1264 unrelaxed Cl 1 Cq: -15.7965 (MHz) Eta: 0.0089 Cl 2 Cq: -16.9778 (MHz) Eta: 0.0048 Cl 3 Cq: -16.9515 (MHz) Eta: 0.0048 Cl 4 Cq: -15.8271 (MHz) Eta: 0.0089 QE EFG DATA (DTN – 35Cl) relaxed Cl 1 Cq= 9.4969 MHz Eta=-0.00000 Cl 2 Cq= -0.9388 MHz Eta= 0.00000 Cl 3 Cq= 9.4869 MHz Eta= 0.00000 Cl 4 Cq= -0.9298 MHz Eta= 0.00000 CASTEP EFG DATA (DTN – 14N) relaxed N 1 Cq: -3.7067 (MHz) Eta: 0.3400 N 2 Cq: -3.3560 (MHz) Eta: 0.4226 N 3 Cq: -3.6918 (MHz) Eta: 0.3463 N 4 Cq: -3.3621 (MHz) Eta: 0.4256 N 5 Cq: -3.6917 (MHz) Eta: 0.3468 N 6 Cq: -3.3533 (MHz) Eta: 0.4231 N 7 Cq: -3.7068 (MHz) Eta: 0.3410 N 8 Cq: -3.3535 (MHz) Eta: 0.4228 N 9 Cq: -3.6920 (MHz) Eta: 0.3466 N 10 Cq: -3.3611 (MHz) Eta: 0.4251 N 11 Cq: -3.7067 (MHz) Eta: 0.3409 N 12 Cq: -3.3524 (MHz) Eta: 0.4230 N 13 Cq: -3.7073 (MHz) Eta: 0.3399 N 14 Cq: -3.3563 (MHz) Eta: 0.4225 N 15 Cq: -3.6927 (MHz) Eta: 0.3469 unrelaxed N 1 Cq: 3.0165 (MHz) Eta: 0.8408 N 2 Cq: -2.6850 (MHz) Eta: 0.9198 N 3 Cq: 3.0154 (MHz) Eta: 0.8423 N 4 Cq: -2.6885 (MHz) Eta: 0.9216 N 5 Cq: 3.0164 (MHz) Eta: 0.8419 N 6 Cq: -2.6841 (MHz) Eta: 0.9213 N 7 Cq: 3.0179 (MHz) Eta: 0.8414 N 8 Cq: -2.6854 (MHz) Eta: 0.9186 N 9 Cq: 3.0188 (MHz) Eta: 0.8421 N 10 Cq: -2.6833 (MHz) Eta: 0.9207 N 11 Cq: 3.0214 (MHz) Eta: 0.8411 N 12 Cq: -2.6802 (MHz) Eta: 0.9177 N 13 Cq: 3.0199 (MHz) Eta: 0.8406 N 14 Cq: -2.6798 (MHz) Eta: 0.9189 N 15 Cq: 3.0198 (MHz) Eta: 0.8417 N 16 Cq: -2.6788 (MHz) Eta: 0.9203 QE EFG DATA (DTN – 14N) relaxed N 45 Cq= -3.5069 MHz eta= 0.36826 N 46 Cq= -3.5084 MHz eta= 0.36870 N 47 Cq= -3.5084 MHz eta= 0.36870 N 48 Cq= -3.1864 MHz eta= 0.38355 N 49 Cq= -3.1864 MHz eta= 0.38355 N 50 Cq= -3.1880 MHz eta= 0.38408 N 51 Cq= -3.1864 MHz eta= 0.38355 N 52 Cq= -3.1864 MHz eta= 0.38355 N 53 Cq= -3.1880 MHz eta= 0.38408 N 54 Cq= -3.5069 MHz eta= 0.36826 N 55 Cq= -3.5084 MHz eta= 0.36870 N 56 Cq= -3.5084 MHz eta= 0.36870 N 57 Cq= -3.5069 MHz eta= 0.36826 N 58 Cq= -3.1880 MHz eta= 0.38408 N 59 Cq= -3.1880 MHz eta= 0.38408 N 60 Cq= -3.5069 MHz eta= 0.36826 Q= 2.04 1e-30 m^2 For NaClO3, CASTEP arrived at an average NQR frequency (over 4 sites) of 28.9 MHz. For chlorine in DTN, CASTEP calculated NQR at 8.5 MHz (unrelaxed) and 4.5 MHz (relaxed). For 14N in DTN, the calculation yields 2.9 MHz and 1.6 MHz for inequivalent sites in the relaxed EFG calculation. The difference between the relaxed and unrelaxed calculation lies in the structural data preparation. In the unrelaxed case, structure data from direct measurements in literature is fed directly into CASTEP and used. In the relaxed case, the structure parameters itself are adjusted using GIPAW, and then the EFG is calculated.