OmniSci Histogram not working

I am trying to do a histogram, to replicate this geospatial analysis:

import ibis

conn = ibis.omniscidb.connect(
    host='metis.mapd.com', user='mapd', password='HyperInteractive',
    port=443, database='mapd', protocol= 'https'
)
t = conn.table("tweets_nov_feb")
x, y = t.goog_x, t.goog_y

WIDTH = 385
HEIGHT = 564
X_DOMAIN = [
        -3650484.1235206556,
        7413325.514451755
      ]
Y_DOMAIN = [
        -5778161.9183506705,
        10471808.487466192
      ]

t[(X_DOMAIN[0] < x) & (x < X_DOMAIN[1])].group_by(
    t.goog_x.histogram(WIDTH).name("x_bin")
).aggregate(t.count()).execute()

However it fails with:

Exception: Exception: Inconsistent return type for FLOOR: SELECT floor((t0."goog_x" - (t1."min_1a6124" - 1e-13)) / ((t1."max_1a6124" - (t1."min_1a6124" - 1e-13)) / 384)) AS x_bin,
       count(*) AS "count"
FROM tweets_nov_feb t0
  JOIN (
    SELECT min("goog_x") AS min_1a6124, max("goog_x") AS max_1a6124
    FROM tweets_nov_feb
  ) t1    ON TRUE
WHERE (t0."goog_x" > -3650484.1235206556) AND
      (t0."goog_x" < 7413325.514451755)
GROUP BY x_bin

[saulshanabrook]

So I think what would be nice on ibis’s side would be to do this cast stuff for me.

[saulshanabrook]

OK fixed that one by specifying binwidth manually:

t[
    (X_DOMAIN[0] < x) & (x < X_DOMAIN[1]) &
    (Y_DOMAIN[0] < y) & (y < Y_DOMAIN[1])
].group_by([
    t.goog_x.histogram(
        base=ibis.literal(X_DOMAIN[0], 'float64').cast('float32'),
        binwidth=ibis.literal((X_DOMAIN[1] -  X_DOMAIN[0]) / WIDTH, 'float64').cast('float32')
    ).name("x_bin"),
    t.goog_y.histogram(
        base=ibis.literal(Y_DOMAIN[0], 'float64').cast('float32'),
        binwidth=ibis.literal((Y_DOMAIN[1] -  Y_DOMAIN[0]) / HEIGHT, 'float64').cast('float32')
    ).name("y_bin")
]).aggregate(
    t.count()
).execute()