Black Lives Matter, Blue Lives Matter: Let's fix this mess

I have stayed silent on this issue because I haven't known what to say.  I didn't know how to begin: how could I contribute anything meaningful to the thousand of voices saying things more eloquently and passionately?  And then I read this today: a letter from a mixed race woman to quiet white friends.  This made me realize that not saying anything is a way of saying something.

So I want to say this: black lives matter and blue lives matter.  The recent (and not-so-recent) killings in both directions are completely unacceptable.  While I haven't been personally impacted by these events, I have loved ones on both sides, as do many people. But they aren't even "sides"; as Trevor Noah said, you can be pro-cop and pro-black.

Something is obviously broken, and I want to do my part to fix it.  For now, maybe all I can do is acknowledge that there is a problem and try to understand it better.  But if everyone did that, we would already have made a lot of progress.

The question isn't whose lives matter.  They all matter.  The real question is: what will you do to make the world better and safer?

Edit/addition: I just wanted to be clear that while "all lives matter," I wasn't trying to dismiss the BLM movement in any way.  BLM focuses on making the world better and safer for black citizens.  This is very necessary, and if successful, they will end up helping the world in more ways than they originally intended.  All I was trying to say is that it doesn't matter who you're trying to help, as long as you are doing something productive and not just dismissing the situation.


the first of the season

Today I harvested the first greens of the season—baby spinach and pea shoots.  The garden has started off lovely.

There are days when I can't believe it's not June yet, and then there are days I feel stuck in March.  But mostly, I just feel tired, and the date doesn't mean anything.

I started teaching seminary in January.  For those unfamiliar, it's a 6am religion class for high school students.  At first, it was very fun, but as time drew on, my excitement turned to exhaustion.  I love deep study of religious texts, but getting up early is not my strong suit.

Combine that with working in the youth program (a second church job), along with something about finishing my PhD in 2016, and you get a stressed, sleep-deprived woman. When I'm not worrying about implementation bugs in my statistical inference algorithm, I'm worried about teens cutting themselves.  It's fun.

I'm trying to find some balance now.  N is helping me so much; one thing he's doing is teaching seminary two days a week.  I need to start doing more things that aren't work and church. Gardening qualifies.


Suicides and "The Policy"

For those of you who aren't aware, the LDS church instituted a policy regarding LGBT families back in November. One of the major criticisms of this policy is it could potentially increase the suicide rate of LGBT Mormons. And recently, some numbers have been thrown around about the number of LGBT Mormon suicides since the policy was instituted.  I haven't fact-checked these, but as of January 23, the numbers I've seen were around 32.

I don't understand the policy, and I don't particularly want to defend it, but I do believe in statistics. The number of suicides alone is not enough to say that the policy has caused any change in suicidal behavior. We need to compare to the number of LGBT Mormons who would have done the same if there were no new policy announced.

We can't actually know what would have happened, but we can estimate it.  Ideally, we'd have data on LDS LGBT suicides rates, but I couldn't find anything quantitative.  So let's assume that most of these cases are Americans, and the average suicide rate in the US is 0.0211% per year, or 0.00451% of the population in the same 78 day period. The church reports around 6,466,267 members in the United States. This means that, if Mormons are typical in their suicide rate (which they seem not to be), there should be around 285 Mormon suicides in that same time period. Around 3.4% of the adult population identify as LGBT, so that takes the estimate down to about 10 suicides.

All these numbers come from Wikipedia, so I'm using questionable sources, but I didn't want to put too much time into this. But with this very loose estimate of suicide rate, it seems there would be a tragic increasea 220% increase.  That seems a little too large to be believable to me.

We could mirror stats on Jewish LGBT youth to help account for the fact that LGBT folks in religious communities are more likely to attempt suicide.  Along these lines, if we assume there are 219,853 LGBT Mormons and that they are four times more likely to commit suicide than national average, that would put our estimate closer to 40 suicides in the same time rangethis would mean that there was actually a decrease in the suicide rate since the policy was enacted.  This may not be as surprising as it sounds, since there was significant backlash against the policy, so LGBT Mormons may have actually received a surge of support from friends and family.

The point I'm trying to make is that it is very, very difficult to make any causal attributions in this case (or in general).  Suicide is incredibly sad regardless of cause, and I also mourn that any person could feel rejected by their religion.  We must take steps to help prevent suicide, and we must be loving and welcoming members of our respective communities.  But part of knowing what steps to take is to get real data on what's happening, not just jumping to the worst conclusions.  Big changes require strong evidence.  And probably lots of hugs in the meanwhile.


HDF5: where have you been all my life?

Nathaniel introduced me to HDF5 around the winter holidays (because, yes, that's the kind of thing we talk about while on vacation), but I just started using in in earnest this past week via h5py.  I may never go back to plain text data storage if I can help itwe'll see if I can convince you too.

This is a script that simulates a big matrix off random data, writes in both formats, and reads from both formats.  The syntax is similar for both.

To compare h5py against plain text, I ran the above (plus timing code added in) 100 times with different random data.  Here are the average results.

plain text
9.44 sec
0.0634 sec
7.94 sec
0.0051 sec
row access
3e-5 sec
6e-4 sec
col access
1e-6 sec
0.016 sec
file size
239 MB
77 MB

If your data is small enough or you need to access almost all of it repeatedly, plain text files might still be good for you.  I usually sample rows from large datasets that eat memory like chocolate cake. So for almost everything I do, h5py is the clear winner.


A letter to potential Trump supporters

Dear potential Trump supporter,

Let me introduce myself: I'm a politically moderate American citizen.  I know that I am moderate because I went to an incredibly liberal college and am a graduate student at a reasonably liberal university.  Meanwhile, the church that I attend is strongly conservative—this means I regularly witness both sides of the political spectrum.  I do not belong to a political party, and all I want (politically) is for the leadership of this country to make the United States of America the best country it can be.

The reason I am writing to you today because we are allies in this.  If you are considering supporting Trump in 2016, you see some appeal in his call to "Make American great again!"

In my opinion, none of the leading presidential candidates are great this year.  Clinton's lack of honesty is particularly upsetting, if unsurprising for a politician.  But I would vote for her over Trump, and I want to explain why.

It all comes down to fear.  I know that the recent terrorist attacks have been frightening, and that the idea of an influx of refugees can be unsettling to some, but we cannot be afraid of change.  The world has always been constantly changing, and the real challenge is not to resist change straight out, but to use it to shape our world for the better. All of our ancestors have felt fear and weathered change—from the men and women of the Revolutionary War to displaced Native Americans, from 1960s Civil Rights Movement activists to the refugees of far too many wars.  Change has shaped our history and shaped us as individuals.  Look back on your life and notice that your proudest moments most likely come from working hard and having the courage to overcome your fears—so too with our country.

My main objection to Trump as a political candidate is that he sensationalizes this fear and uses it to gain media attention and support.  Instead of encouraging us to have courage and find real, lasting solutions, he fans the flame of fear, then claims that he is the one to take it away with proposals that isolate large swaths of the population.  This is not the dynamic I want to have with my President.  I don't mind brutal honesty—in fact, that's one of the things I admire about Christie—but it is not acceptable to isolate citizens and make them feel uncomfortable in their own country.  How would you feel if your religion or ethnicity were singled out in a negative way?  We need a leader who is honest and can solve difficult problems without resorting to pointing fingers at demographic groups, especially not vulnerable ones like recent immigrants and refugees.

Some of my ancestors came over from England before the Revolutionary War and others were more recent immigrants from Poland—but all of my ancestors were immigrants to America.  Some sought religious freedom as they crossed the Atlantic Ocean as Protestants or as they walked across the Great Plains as Mormon pioneers—they were religious refugees in their day.  How can America be great if we treat the modern equivalents of our ancestors with disdain?

So here is my plea to registered Republicans and Democrats alike: don't make me choose between Trump and Clinton.  I would love to see a candidate I can support with confidence.  Better yet, I want to vote with regret, knowing that either candidate would "do our country proud," just in different ways.  But as it stands, it looks like I will vote with a different kind of regret, knowing that neither candidate can push our country toward the excellence it so rightfully deserves.

I don't want to see America revert to some old greatness.  I want it to continually improve, becoming greater and greater with time, just as individuals should strive to be better with each passing day.  That doesn't mean indulging in new ideas without thought, nor does it mean retrenching to some idealized past.  We need to find the path that balances old and new, conservative and liberal, and just and merciful to make America not just great, but exceptional.

A registered voter


Black box variational inference for gammas

Whoa, this has been a long blogging hiatus for me.  I have no excuses other than I've been enjoying life and working hard.  So not excuses, reasons.

I return with a super light-and-fluffy post to share a guide to black box variational inference for gamma-distributed latent variables.  BBVI is very powerful, but I was having trouble applying it to gamma variables, so I asked Rajesh (its creator) for some tips.  I wrote the guide to try out his tricks on a very simple model and share them with other folks that might be having similar issues.  Have fun, ya'll.


the fastest way to embed fonts in a PDF (for me)

I needed to submit a PDF today with all fonts embedded.  I'm working on a Mac with TexShop, and after much angsty Googling, I found lots of answers that seemed to work for other people, but for one reason or another, weren't working for me.

This is what I ended up doing: I created the PDF like usual with TexShop (without all the fonts embedded...I'm looking at you, Helvetica).  Then, I used pdf2ps to convert it to a postscript file, and then Adobe Distiller to convert back to a PDF (with all fonts embedded).  There may be other ps to pdf options that do the font embedding (Dstiller isn't free), but I was sure Distiller did, and I was on a timeline.

This is the PDF equivalent of "just reboot it."  It's silly that embedding fonts is such a messy process.


An Insider's Spain

Way back in May, a week after I submitted a paper (which got in!) and a two days after Nathaniel gave his final oral presentation for his dissertation, we headed off to Spain!  It was a very different sort of vacation, as Nathaniel grew up in Madrid and has an insider's perspective and also lots of friends still living there.

The first day, Nathaniel was actually a little quiet—memories started flooding in as soon as we hit the Madrid airport, and included things as small as being on a particular ramp or hearing the recorded metro voice.  We spent the afternoon walking in the Retiro park, eating bocadillos, looking the book shops, street vendors and performers, and just generally people watching.  Dinner was at Casa Mingo, where Nathaniel's family used to go regularly.

This one is for my mom, the Iris junkie.
The following day, we saw the royal palace, walked through the downtown, saw the museo del Prado, and visited the botanical gardens in Retiro, which had an abundance of irises and other flowers in bloom.  In the evening , we had dinner with a bunch of Nathaniel's friends from school.  It was fun to meet them, joke around, and just have a relaxing evening. Sunday was more memories: we visited Nathaniel's old church congregation, walked by his old apartment and school, and met with another dear friend.

The following day, we left Madrid, which began a very different phase of the trip.  We did some hiking and drove down to Toledo.  There, we saw the El Greco museum, which has both art and living spaces restored to late 16th century style.  We also saw a cathedral, tapestry museum, Jewish museum, and an Alcazar which had been converted to the most elaborate military museum I've ever seen.  Toledo also has a marzipan factory, so I had to put some in my mouth.

Then, off to Córdoba!  After dinner, we had tea and sweets in cute little Moroccan tea shop.  The following dat we saw the Mezquita, which was thoroughly strange.  It was once a mosque, but a cathedral had since been built in the center, and services were going on while we were there.  The guide brochure had a strong catholic bias, with only one of six panels devoted to its Muslim history, most of which read a little bit like "it was a church, and then the Muslims tore that down and built a mosque, and so we feel totally justified in making it a cathedral, because it was once a church.  Do you hear that?  We were here first."  I exaggerate, of course, but it was very defensive.

Mosque-cathedral hybrid. Just a cute alley we found.

We also saw a roman water wheel, and another alcazar, which was a child's dream fort with lovely garden and a fun irrigation system.  I would let you guess who was dorking out about what, but the pictures give it away.

Nathaniel dorking aout about irrigation. Me dorking out about flowers.

Next up was Granada.  We arrived in the evening, caught a flamenco show, and watched the sun set over Alhambra.  We got up early the next day because Alhambra tickets sell out quickly, and that was the major reason we were there.  Thankfully, we got to see it, and it was both amazingly elaborate and extensive.  When I wasn't cooing over the gardens, I was on a mission to find all 17 wallpaper groups, which I'll post about later.  After we spent most of the day in Alhambra, we took a quick hike and then headed over to Huelva, where Nathaniel had another classmate (Charli) we were going to stay with.

Orange and cream colored pomegranate blossoms.

We used Huelva as a launching point to see Sevilla.  There we saw yet another cathedral and alcazar; the latter we recently recognized in a Game of Thrones episode (Dorne).  We also saw an amazing face-off between a duck and a peacock.  The duck was apathetically looking for food while the peacock was aggressively shaking its feathers at the duck.  It went on for about 10 minutes before we moved on.  Eventually we joined up with our hosts Charli and Laura who encouraged us to do silly touristy things.  That evening, it was Charli's goal to present me with food I would refuse.  We began with caracoles (snails), then proceeded to tiny clams, a large octopus tentacle, and finished with little squid.  I ate everything, but by the time I got to the squid, I was very full, so only had one.

Charli's idea.

After departing from Charli and Laura's place the next morning with bellies full of fresh churros dipped in hot chocolate, we made a quick stop at the replicas of Christopher Columbus's ships, and then headed to Segovia.  There, we marveled at the massive Roman aqueduct and had dinner at the famous (both within the Chaney family and without) Restaurante Jose María, which lived up to every expectation.  Jose María himself even came by our table to ask how things were.

For some reason, every statue of Christopher Columbus we
saw has him pointing.

Ordesa was up next—it was a bit of a drive to the park, as it is on the French border.  There we spent three nights up in the mountains.  We saw beautiful wildflowers, and families of chubby marmots and sprightly mountain goats.  The rock formations were simply stunning—the beauty of Ordesa cannot properly be put into words.  I'll just put up a bunch of pictures instead.

Roman bridge.
This waterfall fascinated Nathaniel.
It came out of the side of the rock.

Looking hardcore with my ice pick.
Despite the snow, it was hot.
And because of the snow it was bright.
So we build this shelter for lunch
after deciding an igloo would take too long.
The gap tooth in the background is a pass,
with France on the other side.
His turn to look hardcore.
But I needed the last word in hardcoreitude.

After hiking, we made our way down to Barcelona, which felt far more foreign than I expected, probably due to minor language barriers.  We had a lovely evening with Nathaniel's brother, who was there on business, then saw the old city and the Sagrada Familia the following day.  At first I was hesitant to go inside the basilica, as it reminded me of a drippy sand castle, but inside, the architectural mastery was shown to its fullest, as the colored glass windows cast fantastic shadows over smooth geometric shapes.  The day after, we saw Castell de Montjuïc and the area surrounding the Olympic grounds.

Like any long trip, I was ready for the routine of home at the end, but I was less exhausted than trips usually make me.  I think this was due to our slower-than-usual pace and much of it being already familiar to Nathaniel.  If we are blessed with the ability to do so, I would love to make Spain a regular vacation destination for us, returning every five or six years.  It has a piece of home.

Let's end with this awesome flower we saw in Barcelona.
Two petals are each half pink and half white.