hgrecco / pint / 940

91%

push

travis-ci

Merge #703

703: Fix dequantify bug r=hgrecco a=znicholls

At the moment, if you take a pint dataframe and dequantify it, you lose all the index information e.g. 

```python
df
```

```python
    speed mech power torque  rail pressure fuel flow rate fluid power
      rpm         kW    N m            bar          l/min          kW
a c  1000        NaN     10           1000             10         NaN
b c  1100        NaN     10  1000000000000             10         NaN
a d  1200        NaN     10           1000             10         NaN
b d  1200        NaN     10           1000             10         NaN
```

`quantify` works fine

```python
df.pint.quantify(ureg)
```


```python
      speed mech power torque    rail pressure fuel flow rate fluid power
a c  1000.0        nan   10.0           1000.0           10.0         nan
b c  1100.0        nan   10.0  1000000000000.0           10.0         nan
a d  1200.0        nan   10.0           1000.0           10.0         nan
b d  1200.0        nan   10.0           1000.0           10.0         nan
```

`dequantify` after quantification doesn't (and also changes the type of nan oddly enough...)

```python
df.pint.quantify(ureg).pint.dequantify()
```

```python
                   speed mech power         torque rail pressure fuel flow rate fluid power
  revolutions_per_minute   kilowatt meter * newton           bar liter / minute    kilowatt
0                 1000.0        NaN           10.0  1.000000e+03           10.0         NaN
1                 1100.0        NaN           10.0  1.000000e+12           10.0         NaN
2                 1200.0        NaN           10.0  1.000000e+03           10.0         NaN
3                 1200.0        NaN           10.0  1.000000e+03           10.0         NaN
```

This is my attempt to fix that. @andrewgsavage might have a better solution/suggestions of other tests we should write.

Annoyingly I had to make our test cs... (continued)

7 of 7 new or added lines in 1 file covered. (100.0%)

2872 of 3247 relevant lines covered (88.45%)

11.08 hits per line